September DECUServe Journal
fr...@eisner.decus.org
=====================
September, 1993
Hello! This issue of the DECUServe Journal contains many interesting
articles. We have some information on HSC's, VMS internals, a discussion of
VMS database packages, Macintosh network backups, and disk encryption. And
then you can cool your heels by reading a fun discussion about telemarketing.
Enjoy!
Email (and questionaire responses) to:
Sharon Frey
DECUServe Journal Editor
fr...@eisner.decus.org
Title of Contents
-----------------
Editor's Comments ....................................... 1
Technical Information ................................... 40
Articles
So much for HSC Redundancy! .......................... 2
Justifying Internet .................................. 9
Is high kernel mode time common? ..................... 15
In Search Of The Perfect Database Package For VMS! ... 18
Macintosh network backups? ........................... 26
The Art and Science of Telemarketer Baiting .......... 27
Encrypting a disk at the physical level .............. 36
September, 1993 DECUServe Journal Page 2
So much for HSC Redundancy!
---------------------------
This article is an extract of the DECUServe Hardware_Help
conference topic 1335. This discussion occurred between
February 4, 1993 and February 14, 1993.
By David Johnson, Jorma Tapola, Terry Kennedy, Dale Coy, Keith Chadwick,
Brian Tillman, Carl Friedberg, Gus Altobello, George Cornelius
(02/04/93 Johnson)
------------------
We had a HSC50 die on us about 2 PM today. That's not so bad because all
of the drives are dual ported. However, when the HSC50 died it corrupted 5
drives. We have spent the last 5 hours restoring data and verifying other
drives on the HSC were valid. Field Service has not found any problems yet!
They think it was caused by the m.data module or the power supply. Needless
to say, it's out of here!
So much for dual porting and shadowing.
(02/04/93 Tapola: What are disk types?)
----------------------------------------
I'm curious. We've had many HSC crashes without losing data.
What type of disks are you using? Or are 100% sure it's HSC?
(02/04/93 Kennedy: What sort of corruption?)
---------------------------------------------
Were these DEC drives, or 3rd-party "RA lookalikes"? There is a lot of
error checking built into the SDI protocol and I find it impossible to believe
that multiple DEC drive were corrupted by a SDI-related fault (I can believe
that one or more drives on a single requestor were affected, though).
However, there could be some subtle corruption introduced by a software
problem, or HSC software corruption introduced by a failing module or power
supply in the HSC, although a lot of that has extensive error checking as well.
I think we need to define what you mean by "corrupted" - was the data on
the drives just changed improperly (like writes went to the wrong blocks or
drives) or was there a loss of format or format inconsistency (the HSC VERIFY
utility should check this for you)?
(02/04/93 Coy: Thanks)
-----------------------
These were 5 shadow sets? With the members on different requesters but on
the same HSC? And Phase I shadowing?
Very curious. Of course, there is *NO* perfect hardware and/or software.
There will always be _some_ probability that a hardware failure in an HSC (or
some combination of multiple failures)
September, 1993 DECUServe Journal Page 3
However, my long-term experience is that I have _never_ seen an HSC corrupt
anything on any disk. You may have been terribly unlucky. If so - thanks for
soaking up all of this century's log-probability failures. :-)
(02/04/93 Johnson: RA70s and RA90s)
------------------------------------
>We've had many HSC crashes without losing data.
So have we. We have also been using HSC's for 6 years without anything
like this happening.
>What type of disks are you using? Or are 100% sure it's HSC?
DEC RA70S and RA90. 5 drives were corrupted that were all mounted on the
same HSC. Yes, 100% sure.
(02/04/93 Coy: How about the shadowing?)
-----------------------------------------
Let me ask this again, then.
>> So much for dual porting and shadowing.
> These were 5 shadow sets? With the members on different requesters but
> on the same HSC? And Phase I shadowing?
Also, I understand how you feel. However,
Without dual porting, you can't do any failover. Is that better?
Without shadowing, you only have one copy. Is one better than two?
OK - let's say you had these 5 drives on a separate controller, with no
shadowing. And the controller failed and trashed the 5 drives.
The HSC doesn't seem to have made things worse for you. And it has
(probably) saved you a few times over the past 5 years. It's just not
absolutely and totally perfect.
(02/04/93 Johnson: More Details...)
------------------------------------
More details...
I have had some sleep now so I will give you all more information.
Sometime yesterday afternoon, something happened to one of out production
HSC 50s. It never actually crashed. However, five out of the ten or so drives
that were mounted through it were corrupted. When the corruption occurred, the
drives tried to unsuccessfully fail over to another HSC.
Two of drives were a shadow set on different requester cards. After this
happened, these two drives were totally inaccessible from any member in the
cluster. When you tried to mount them a Fatal Controller Error message was
returned. This error occurred even if you tried to mount them from another
HSC. The HSC DKUTL utility indicated that all four copied of the RCT was
corrupt on both drives. We had to reformat the drives and restore from the
latest backup.
September, 1993 DECUServe Journal Page 4
Another RA90 had the same RCT corruption problem. However, this drive's
shadow set was not effected at all.
Another shadow set of 2 RA92s had a few blocks corrupted but was still
accessible from the cluster. However, when an application would access these
blocks, a CLASS_DRIVER bugcheck would crash the node. At the same time, the
HSC serving the disk would produce an EDC error. We reformatted one of the
shadow set members and restored it from the latest backup. We still have the
other one off line. When I get a chance, I will run the HSC Verify utility on
it and see what it says.
All other drives that the HSC was serving at the time of the problem were
not effected. The drives that were effected were on different requesters. Also
note that we are running controller based shadowing. The current theory is
that power supply or the M.DATA module had a problem and that all the write
operations in progress were corrupted.
Don't get the wrong impression from this note. I think we are the victims
of a very rare HSC problem. Field Service told me that they have seen this
happen before. I think HSCs and shadowing are a good way to configure for
reliability. However, we are going to start looking at host based shadowing.
We have been using HSCs for over 6 years and we are currently running 8 of them.
This is the first time anything like this has happened!
(02/04/93 Coy: A rare event)
-----------------------------
Host based shadowing might not have helped much in this situation. You
_can_ configure host-based shadowing so that the shadow set drives are on
totally different HSCs. In the unlikely event that you had, this would have
saved a disk drive. However, you give up some big points in efficiency of
merge-copy operations.
I just switched our system from Phase I to Host-Based shadowing, mostly
for performance reasons. I am very happy with the results.
(02/05/93 Johnson)
------------------
> However, you give up some big points in efficiency of merge-copy operations.
Why, is the merge-copy performed differently when both shadow sets are on
the same HSC?
What kind of other performance gains did you see?
(02/05/93 Coy: HSC hardware assist)
------------------------------------
Writing from "fuzzy memory", yes. There is "hardware assist" for the
Phase II shadowing in (at least some of the) HSCs. At the most recent level of
HSC software. I'm sure this applies to HSC V650.
There is this distinction between "copy" and "merge" operations, that I
always get mixed up in my memory, and have to go check the book.
But I am certain about the "hardware assist" for some Phase II operations,
and (obviously) it only applies if the disks are on the same HSC.
September, 1993 DECUServe Journal Page 5
(02/05/93 Chadwick: VMS V5.5-2 is also a requirement)
------------------------------------------------------
You must also be running VMS V5.5-2 to have the shadowing software take full
advantage of the HSC hardware assist for phase II shadowsets on the same HSC.
Refer to section 2.30 in the VMS Version V55.-2 Release Notes (pages 2-33 ]
through 2-43).
(02/05/93 Coy: Uh-Huh)
-----------------------
Thanks for pointing that out. It's an important point, and I forgot to
mention it.
We just went to 5.5-2, installed a 7000-610 into the cluster, installed
8 more disks on our pair of HSCs, re-volume-named all of the disks (old and
new), and converted from Phase I to Phase II Shadowing. All in one week. [And
now you know why I can't seem to remember which things were prerequisites for
which other things]
To reiterate - I am very happy with the performance of Phase II
Shadowing. In our situation, it is performing noticeably better than Phase I
Shadowing in some respects - and there is no place that it is worse.
(02/06/93 Tillman: Phase II costs a lot more than Phase I)
-----------------------------------------------------------
I'd love to switch to Phase II shadowing, but it would be much too costly
to go out and buy licenses for the 30 workstations in our cluster that don't
have disks. As it is, we have shadowing licensed on the five CI nodes and they
serve the disks to the remainder of the cluster (those 30 other nodes).
That's one thing I have always wondered about. The *real* I/O requests
are still being performed by the CI nodes serving the disks to the NI nodes.
Why would the NI nodes need any shadowing software?
(02/06/93 Coy: If you don't need the MOUNT command...)
-------------------------------------------------------
See VMS Topic 1515. The basic reason is that Mounting a Phase II Shadowed
Disk requires it. [Now, maybe it _shouldn't_, but that's another question].
BTW - Note that there is special discounted pricing on Phase II Shadowing
licenses through the end of March 1993.
(02/07/93 Tapola: Backups are usefull!)
----------------------------------------
I'll like to point out - no doubt that everybody reading this thread knows -
that backups should be taken care of, even with any kind of hardware.
My fear is that RAID is going to raise questions like:"Why do we need
backups ?"
Yes, we DO NEED backups?
September, 1993 DECUServe Journal Page 6
(02/07/93 Coy: Of course)
--------------------------
I always answer with a question, simply -
What is the effect on your system of doing
DELETE [...]*.*;*
I have yet to find someone who continues to believe that Shadowing or
RAID N will eliminate the need for backups.
(02/07/93 Chadwick: Would you buy RAID from a used car salesman?)
------------------------------------------------------------------
As Dale has already poijnted out, RAID will cheerfully delete all portions
of the file when requested to do so by a suitably privilged user.
Another thing to watch out for - All the claims for RAID reliability are
based on calculations which assume uncorrelated failures.
A power failure/glitch and disks failing to spin back or the file system
being trashed by noise is classified as a CORRELATED failure.
The RAID vendors get very closedmouthed when you ask them about real world
experience and failure rates with their products...
(02/11/93 Friedberg: questions about HSC phase II shadowing)
-------------------------------------------------------------
I have no HSC's to play with these days (aside from the code inside the
RF73's) but I am curious why you would want to have members of a phase II (host
based) shadow set in the SAME HSC?
(1) Availability considerations would suggest you benefit by keeping
members of the shadow set on different HSCs
(2) I don't recall the exact details of the restrictions onf 5.5-2
minimerge, but on the RF73's, it relies on the cache within the drive
(read, HSC) to keep a write-log, and thus can figure out which disk
blocks might be different on different shadow-set members. Naively, I
would guess (...) that mini-merge would also work on suitable HSC's (ie
meeting the stated minimum hardware/software configurations), with
members in different HSC's. I HOPE it works that way. COmments?
(3) Performance: Yes, you would avoid the CI paths, etc., during full
copies, but how often do full copies get done, and how much difference
is there? Mini-merge is supposed to change the hours into seconds
(haven't had it happen yet, as we NEVER crash a node...)
(4) I think there is VMS thread somewhere on satellite-nodes
destroying shadow sets (which I posted over a year ago in a
SCSI-dual-host topic). Yes, every node (satellite or direct-connect)
masters each shadow set, and any node can remove a member from the
shadow set if it loses touch with that member. I can't believe they did
that, especially since VMS engineers must work in clusters with LOTS of
September, 1993 DECUServe Journal Page 7
satellites constantly booting in and out. All it takes is a brief
fluctuation on the ethernet, and you loose a member from your shadow
set (at least, you start a full copy in that case...) I hope this is
fixed "in a future release?"
(02/11/93 Kennedy: HSC's only know about VAXen, not other HSC's)
-----------------------------------------------------------------
>Naively, I would guess (...) that mini-merge would also work on suitable HSC's
>(ie meeting the stated minimum hardware/software configurations), with members
>in different HSC's. I HOPE it works that way. COmments?
But HSC's don't maintain connections to each other - they don't even know
the other one is "out there". So the mini-merge needs to go HSC1 -> VAX ->
HSC2. The VAX can't even create a direct HSC1-HSC2 path.
(02/11/93 Coy: It's the "no free lunch" theory)
------------------------------------------------
>RF73's) but I am curious why you would want to have members of a phase II
>(host based) shadow set in the SAME HSC?
> (1) Availability considerations would suggest you benefit by keeping
>members of the shadow set on different HSCs
Exactly.
It is important to understand the availability issues with various
configurations (2 HSCs, same HSC and same/different requestor), as well as the
performance issues (merge, seek optimization, etc) and the cost issues.
If there were a single answer, we wouldn't be discussing it. In general,
performance can be better if the drives are all on the same HSC. But if the HSC
fails, performance goes to zero.
(02/11/93 Altobello: Has HSC failover gone away in Volume Shadowing Phase II?)
-------------------------------------------------------------------------------
>> (1) Availability considerations would suggest you benefit by keeping
>> members of the shadow set on different HSCs
>But if the HSC fails, performance goes to zero.
I'm only running Phase I shadowing, but surely this isn't true? If a pair
of disks is dual-ported to HSC000 and HSC001, but both are being served by
HSC000 (to get the performance boost) and HSC000 fails, won't HSC001 pick them
up and continue running?
Where is the "zero" in all this? At worst, you'd get a full copy when the
new HSC picked the drives up.
Considering that I shadow because "drives fail", and backups aren't perfect,
and downtime is expensive (while one replaces the drive and restores the
backup); considering the performance boost you're reputed to get with Volume
Shadowing Phase II on HSCs with VMS v5.5-2; and considering how seldom HSCs
fail: Why doesn't it make sense to have the drives on a single HSC.
Remember, we're talking "high availability" here, not "high reliability" --
none of this hardware provides the latter.
September, 1993 DECUServe Journal Page 8
(02/11/93 Coy: Thanks for clarifying what I tried to say)
----------------------------------------------------------
>> (1) Availability considerations would suggest you benefit by keeping
>> members of the shadow set on different HSCs
>But if the HSC fails, performance goes to zero.
Uh - please don't quote me out of context. Those two things were separated
by several lines of text. I was just trying (poorly, it turns out) to point out
that it is sometimes necessary to balance conflicting objectives, when working
with HSCs.
> I'm only running Phase I shadowing, but surely this isn't true? If a pair
>of disks is dual-ported to HSC000 and HSC001, but both are being served by
>HSC000 (to get the performance boost) and HSC000 fails, won't HSC001 pick them
>up and continue running?
It sure will. The "availability considerations" are mostly the oddball
ones (not unique to Phase I) such as "how do you know the other HSC is OK" and
similar low-probability events.
> considering how seldom HSCs fail: Why doesn't it make sense to have the
>drives on a single HSC.
That's the way I have mine set up. The shadowed drives are on different
requesters in each HSC, and they are normally "ONLINE" to the same HSC. And
they fail over just fine.
OTOH, there are other configurations that might suit special problems.
For instance, there's mention around here of a (low-probability) HSC failure
that trashed both members of a shadow set. If that worried me a lot, I might
give up the performance boost.
(02/14/93 Cornelius: I would think mini merge would work..)
------------------------------------------------------------
> But HSC's don't maintain connections to each other - they don't even know
>the other one is "out there". So the mini-merge needs to go HSC1 -> VAX ->
>HSC2. The VAX can't even create a direct HSC1-HSC2 path.
Yes, that's true if a mini-merge is a full merge. But isn't the idea of a
mini-merge the following:
1. Node crashes;
2. Second node asks controllers to return information concerning
operations that were in progress (from crashed node? from
all nodes?) to either shadow member at time of crash;
3. Second node performs appropriate block copy operations to
make sure that the data for any blocks being written
by crashed node is the same for both shadow members.
This being the case, all that is required is that the controllers have the
proper level of support, and that the "second" node have a path to both
controllers.
September, 1993 DECUServe Journal Page 9
Now if the question has to do with full shadow copies being controller-
assisted, I would agree - it's hard to have controller-assisted shadow copies
if the two HSC's cannot talk to each other.
(02/14/93 Coy: Is this RAID 23?)
---------------------------------
Well, I sure thought that I said that you needed to have the disks on the
same HSC, in order to have HSC-assisted operations.
My feeble old brain can't keep straight what's a "mini-merge" and what's a
"merge-copy" and what's a "shadow-copy" and what's a
"maxi-merge-shadow-mini-copy".
Justifying Internet
-------------------
The following article is an extract of the DECUServe
Internet conference topic 140. This conversation occurred
between April 27, 1993 and June 4, 1993.
By Curt Snyder, Matt Holdrege, Terry Kennedy, Don Roberts, Tom McIntyre,
Billy Youdelman, Tom Allebrandi
(04/27/93 Snyder)
-----------------
I work for a Fortune 500 pharmaceutical company, which does NOT have
Internet access (probably one of the last). I am in the process of justifying
the connection to mgmt. I am looking for benefits of Internet access for both
scientific and business areas. Hard numbers (dollars) are gratefully
appreciated, but intangibles would be good also. Our current estimate is that a
full time connection would cost about $24K in hard money annually. What other
costs are buried in this (assuming hardware is already in place and we have
system manglers)?
(04/27/93 Holdrege: A good deal all around)
--------------------------------------------
Well there are enormous justifications for scientific research. Most of the
rest of the worlds researchers (in any discipline) use the Internet to
distribute information and conference.
Speaking just for MIS research and professional development, the knowledge
you gain from using the Internet will pay for itself year-in, year-out.
I'm surprised at your $24K annual cost estimate. Our 56K link from PSI costs
about $7400 (connection fee plus circuit costs.) We also had to put up a few
grand to buy the router & CSU/DSU. Even if we doubled our connection fee it
would only be $14,400 per year.
And best of all, you would have a free connection to DECUServe! 8-)
September, 1993 DECUServe Journal Page 10
(04/28/93 Kennedy: Shop around)
--------------------------------
PSI's pricing is very strange. Here in NJ (next to their original "home
base") they were quoting $21K/year for "access", with hardware and line costs
extra. Also, their performance in this neck of the woods is pretty abysmal -
before they recently upgraded, you took a 160ms "hit" crossing their NSF
gateway. And they were not responsive to complaints - an educational site which
has multiple T1's was getting unsatisfactory service. About a year ago I made
an informal report to the NIC and PSI that they were not complying with the
requirements for registering domains, and beyond a "we'll think about it",
they never responded. [BTW, the issues were that they would not list an
organizations full name if the "common name" was an acronymn - so if we were a
PSI customer our WHOIS record would say "SPC" followed by our address, rather
than "SPC" "St. Peter's College" and the address, and that they always
registered their own two people as admin and tech contact, never updating them
when they changed people, leaving no way to send mail to the responsible people
at the end site, and that they do not track their customer's relocation. In one
case, a PSI customer moved several years ago to another state (including their
56Kb line, so PSI can't say they "didn't know"), yet the WHOIS entry still
lists the old address.]
Anyway, the best advice is to shop around. In my opinion, a somewhat higher
price is justified by better support and reliability. After some initial
problems with my local provider (JvNCnet) we've been very happy with their
responsiveness.
On general principles, I'd pick UUnet/AlterNet over PSI because of good
experiences with the former and poor with the latter.
Oh, and we pay something like $12K/year for "membership" which includes the
router and maintenance on it, and about $7K/year for the 56Kb line. Of course,
these numbers are old since it's a 3-year NSF grant - current pricing may be
higher.
(04/28.93 Roberts: You can't justify NOT having a connection)
--------------------------------------------------------------
Beckman Instruments, where I work, is in the medical instrument business.
In some ways you're probably a competitor, but, hey, this is DECUS, right?
We have found a great deal of benefit for us in several areas.
First, there is a hierarchy called bionet. This has all the cutting
edge information on biology and biological sciences. It is also the media of
choice for distributing new information and papers amongst the academic
biological science community. Our scientists and chemists make extensive use of
the data available, frequently downloading experiment results and saving
themselves a lot of work. We also keep in touch with new findings in the field
which helps us stay competitive and develop new products. This by itself has
paid for our connection.
Another area is contacting experts in several fields such as clinical
chemistry where we are trying to stay in the cutting edge. We have a constant
flow of electronic mail to scientists at university hospitals and exchange
information.
September, 1993 DECUServe Journal Page 11
By being connected, we are building good will in our current and potential
customer base. Those amongst our customers who are connected find the quick
access to our scientists to be a good source of new information about our
products and a quick way to get support. Our customers now have access to our
developers and don't have to go through service for many scientific questions.
Also, OUR staff gets to feel like they are still connected to their
academic roots, making them feel less isolated from the sources of their
training and more comfortable. This means our scientists don't feel like
they've left the comfortable college atmosphere and are much happier on the job.
Last, we've been solicited for new business many times via our
connection, so we've (inadvertantly) added a new avenue of marketing (which we
don't actively pursue, but it's still happening).
(04/28/93 Holdrege: Every vendor has their problems)
-----------------------------------------------------
I have heard a lot about PSI's problems in the past. The one's that Terry
mentions are old, I believe. I am registered as the tech contact and they did
spell out our full name at the NIC. They also fixed an overloaded (over-sold?)
router in Santa Clara when I complained about it crashing occasionally. But
every vendor has this problem.
Nearnet is routing DECUS traffic around a microwave ring because of a cisco
bug in one of their routers. This causes packets destined for DECUS to go
through a couple extra routers and lots of distance.
PSI seems to offer the same basic service at half the price. Of course
prices can change.
(04/28/93 McIntyre: Sometimes the line charges can be significant)
-------------------------------------------------------------------
Another issue that does not impact a full linkage but does affect us poor
folk in the boonies is the toll charges for dial up connections. We are
currently using uunet, but discovered from a recent posting on DECUServe that
psi is a local call for us instead of the 800 surchage we pay for uunet.
(04/29/93 Kennedy)
------------------
> I have heard a lot about PSI's problems in the past. The one's that
>Terry mentions are old, I believe. I am registered as the tech contact and they
>did spell out our full name at the NIC. They also fixed an overloaded
>(over-sold?) router in Santa Clara when I complained about it crashing
>occasionally. But every vendor has this problem.
True, but according to your whois, your admin contact is sch...@psi.com.
The delay problem seems to be corrected in the NY area, but there is still a
problem where the first packet seems to wander around for a while - it may be
that there isn't enough memory in the routers to store the route information
(something which will become more and more common all over, I think).
September, 1993 DECUServe Journal Page 12
(04/29/93 Youdelman)
--------------------
> They also fixed an overloaded (over-sold?) router in Santa Clara when I
>complained about it crashing occasionally. But every vendor has this problem.
This was done a couple weeks ago I believe, and it did make a noticable
difference. In general, over the last year or so, PSI's performance has
improved to where the response time is now usually quite good. In testing
the RT-11 Kermit by copying back and forth from here (3000 blocks) over
the net, it's been a while since I've seen it time out or even slow down
much..
(04/29/93 Snyder: Not a competitor)
------------------------------------
Don,
Thanks for the info. FYI, we aren't a competitor with Beckman (remember back
when both Allergan (my employer) and Beckman were part of SKB?). I'm also just
down the road from you in Irvine. We are considering using CerfNet for our
connection also. Any comments on their support, etc? I assume you are
connected to UCI.
A general question. Our local connection offers three different speed
connections for full time: 14.4 KBS, 56 KBS and 1544 KBS. I don't expect a REAL
heavy load for a while, probably some e-mail and a few megabyte downloads on a
daily to weekly basis. I realize faster is better, but can anyone give me a
feeling for which of these they have found adequate? Or recommendations, etc?
The slower the cheaper and I still have to go beg for money from the labs, but
if it's too slow it becomes useless.
(04/29/93 Kennedy: I'd take the 56Kb)
--------------------------------------
I'd start with at least the 56Kb service level. It works pretty welll, and
you can get 6-7Kbyte/second for FTP transfers (assuming the other end is on
an equal or faster line).
You may have options between 56Kb and 1.5Mb - while those are the only line
speeds the phone company will give you, you can often get an in-between speed
like 256Kb by having your vendor speed-limit your port on their end. This
runs over a 1.5Mb line, giving you low latency and better performance. BTW,
this is what DECUS does. Also, upgrades in that space (going from 128Kb all
the way to 1.5Mb) are all "soft", done in the router. No phone line change is
required.
In any event, unless you forsee a lot of traffic, you'd probably start at
the 56Kb level - if your provider offers a choice of routers and interface
gear, you might want to get one that will support faster lines (some don't).
September, 1993 DECUServe Journal Page 13
(04/29/93 Roberts: CERFnet)
----------------------------
> Don,
> Thanks for the info. FYI, we aren't a competitor with Beckman (remember
>back when both Allergan (my employer) and Beckman were part of SKB?).
That was then, this is now :-)
>I'm also just down the road from you in Irvine.
Gee, maybe we should do lunch sometime. Let's grab Matt Holdredge,
Harrison Spain, J.M. Ivler and a few other OC DECUServers and make an event of
it :-)
>considering using CerfNet for our connection also. Any comments on their
>support, etc? I assume you are connected to UCI.
CERFnet gives excellent service, and UCI is very cooperative.
>A general question. Our local connection offers three different speed
>connections for full time: 14.4 KBS, 56 KBS and 1544 KBS. I don't expect a
>REAL heavy load for a while, probably some e-mail and a few megabyte downloads
>on a daily to weekly basis. I realize faster is
We started with the 14.4 route for the same reasons you cite. Don't do
it. One FTP download and you'll know why :-) We're at 56K now and its
sufficient for a full newsfeed and all the FTP and TELNET connections our
company has been interested in so far.
(04/30/93 Holdrege)
-------------------
> True, but according to your whois, your admin contact is sch...@psi.com.
Yes PSI wanted to hang on to the adminstrative contact and I didn't see
the need to fight them about it. One thing for anyone to consider, the IP
address blocks that the service providers assign to you are owned by the vendor.
If you switch your service with PSI, Cerfnet or whomever, you have to give up
your IP addresses.
I think that the problem with PSI's core routers was CPU rather than
memory. The routers can be fitted with loads of memory, but the utilization is
increasing so much that the 68xxx chips in some core routers are maxed out.
(04/30/93 Kennedy: Eh?)
------------------------
Eh? We had our IP address before we selected a network provider. The
addresses are assigned by the NIC (Internic, this week). Your provider may
choose to handle your paperwork for you, but in almost all cases they're
a unique-to-you assignment. The only example I know of that's otherwise is
Barton Bruce's CENTnet (which [at least used to] assigns subnets from the
main CENTnet assignment). However, back then CENTnet was more of a "coop"
than a regional.
September, 1993 DECUServe Journal Page 14
And this shouldn't matter except for the few MILnet hosts who haven't
heard that host tables went out of style years ago. Sure, there are some folks
with broken/missing resolvers, but is that *your* problem?
(04/30/93 Roberts: I agree with Terry)
---------------------------------------
> Yes PSI wanted to hang on to the adminstrative contact and I didn't see
> the need to fight them about it. One thing for anyone to consider, the
Matt, I think you're too easy. We have our own class b address which we
registered on our own. CERFnet never gave us a peep.
As I understand it, PSI has a class b license from which they issue ranges
to their customers, but I was not aware they refused to carry customers with
their own IP addresses. I was under the impression the ranges they issue were
for customers that either didn't have or didn't want to bother with their own
address.
(04/30/93 Holdrege: I look into this further)
----------------------------------------------
Sorry, I didn't mean to imply that PSI or anyone else wouldn't carry
someone who already had thier own address. I just thought that PSI "owned" a
block of addresses that they assigned to their customers who needed them.
Here is a sample of listing from the NIC.
PSI (NET-PSINET-B-10) PSINET-B-10 149.10.0.0
PSI (NET-PSINET-B-100) PSINET-B-100 149.100.0.0
PSI (NET-PSINET-B-101) PSINET-B-101 149.101.0.0
PSI (NET-PSINET-B-102) PSINET-B-102 149.102.0.0
PSI (NET-PSINET-B-103) PSINET-B-103 149.103.0.0
PSI (NET-PSINET-B-104) PSINET-B-104 149.104.0.0
PSI (NET-PSINET-B-105) PSINET-B-105 149.105.0.0
PSI (NET-PSINET-B-106) PSINET-B-106 149.106.0.0
PSI (NET-PSINET-B-107) PSINET-B-107 149.107.0.0
PSI (NET-PSINET-B-108) PSINET-B-108 149.108.0.0
PSI (NET-PSINET-B-109) PSINET-B-109 149.109.0.0
SURFnet BV (NET-SURFNET-BLOCKB77) SURFNET-BLOCKB77 145.81.0.0
SURFnet BV (NET-SURFNET-BLOCKB78) SURFNET-BLOCKB78 145.82.0.0
SURFnet BV (NET-SURFNET-BLOCKB79) SURFNET-BLOCKB79 145.83.0.0
SURFnet BV (NET-SURFNET-BLOCKB80) SURFNET-BLOCKB80 145.84.0.0
SURFnet BV (NET-SURFNET-BLOCKB81) SURFNET-BLOCKB81 145.85.0.0
SURFnet BV (NET-SURFNET-BLOCKB82) SURFNET-BLOCKB82 145.86.0.0
SURFnet BV (NET-SURFNET-BLOCKB83) SURFNET-BLOCKB83 145.87.0.0
UUNET Technologies (NET-UUNETCUSTB32) UUNETCUSTB32 153.32.0.0
UUNET Technologies (NET-UUNETCUSTB35) UUNETCUSTB35 153.35.0.0
UUNET Technologies (NET-UUNETCUSTB36) UUNETCUSTB36 153.36.0.0
UUNET Technologies (NET-UUNETCUSTB37) UUNETCUSTB37 153.37.0.0
UUNET Technologies (NET-UUNETCUSTB38) UUNETCUSTB38 153.38.0.0
UUNET Technologies (NET-UUNETCUSTB39) UUNETCUSTB39 153.39.0.0
September, 1993 DECUServe Journal Page 15
Advanced Network & Services, Inc. (NET-ANS-BNET3) ANS-BNET3 152.158.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET6) ANS-BNET6 152.161.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET7) ANS-BNET7 152.162.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET8) ANS-BNET8 152.163.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET9) ANS-BNET9 152.164.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET10) ANS-BNET10 152.165.0.0
Advanced Network & Services, Inc. (NET-ANS-BNET11) ANS-BNET11 152.166.0.0
(05/01/93 Youdelman)
--------------------
> We started with the 14.4 route for the same reasons you cite. Don't do
>it. One FTP download and you'll know why :-)
Just in case you have to use 14.4, it's a lot cheaper via dial-up (which
I'm pretty sure CERFnet offers) and works just as well. However, it's a one
thing at a time situation if you want decent response time.
(05/01/93 Roberts: This service is called "Dial-n-Cerf" :-) )
--------------------------------------------------------------
(05/02/93 McMahon: CERFnet Dial-Up Bah?)
-----------------------------------------
TGV's San Diego office currently uses Dialup to CERFnet. My understanding
is that they have been very unhappy with the service, and are pursuing other
alternatives.
More details as I hear them.
(06/04/93 Allebrandi: 14.4 may be ok)
--------------------------------------
One way to get away with 14.4 is to see if the provider will do it the way
Alternet does it. We got a 3002 line to Alternet and hung a Telebit T3000 on it.
The T3000 is setup for V.32bis (14.4kb) and V.42bis (error control and
[theoretical] 4:1 compression) for a [theoretical] 57.6kb.
Depedending on where I am going I can get 3200 bytes/second on an FTP. Our
average FTP seems to run around 1100 but that is also with a news feed
up and a couple of Telnets.
Is high kernel mode time common?
--------------------------------
The following article is an extract of the DECUServe
All-In-1 conference topic 911. This discussion occurred
between July 13, 1993 and July 26, 1993.
By Pat Scopelliti, Don Vickers, Dale Coy, Lynda Peach, Matt Holdrege,
Steve Bonine
September, 1993 DECUServe Journal Page 16
(07/13/93 Scopelliti)
---------------------
We're running what seems like a fairly 'stock' ALL-IN-1 system:
Two VAX 4000-600s DSSI clustered running ALL-IN-1, pathworks, and some
other minor stuff (RJE, SNAte, etc.). ALL-IN-1 has been modified (usual stuff,
scripts, forms, etc.). VMS V5.5-2, ALL-IN-1 V3.0, about 600 concurrent
ALL-IN-1 users across the cluster.
We typically measure kernel mode time at 50% of all CPU time consumed
(DECps). This seems to me to be rather high, but until I get DECps V1.1 I
can't tell where this kernel mode time is being spent.
Can some other ALL-IN-1 sites confirm/deny that this is typical of ALL-IN-1?
(This note ties in with topic 332 in the 3RD-PARTY conference which
deals with a product called Memory Master and a search to reduce kernel mode
time).
(07/14/93 Vickers: It's high - a couple of possible sources)
-------------------------------------------------------------
50% kernel mode is certainly on the high side. I have seen it this high on
some ALL-IN-1 systems. In the cases I recall, the sources were either page
faulting or usage of the ALL-IN-1 subprocess.
In one case, the working set sizes (WSQUO and WSEXTENT) were too small which
resulted in higher page fault rates. Obviously, ALL-IN-1 page faults a lot are
page faults occur in kernel mode.
The ALL-IN-1 subprocess also entails a lot of kernel mode due to the number
of context switching that is required. The structure of the ALL-IN-1 subprocess
is such that there are two context switches per line of DCL executed. This adds
up in BIG hurry.
This is two POSSIBLE causes of high kernel mode usage. There are
certainly MANY more lurking out there.
The 'typical' ALL-IN-1 system should be using less than 30% of kernel mode
time as I recall. It will be interesting to see what other folks see on their
systems.
Hope this helps,
(07/14/93 Coy: As usual - it all depends
-----------------------------------------
> We typically measure kernel mode time at 50% of all CPU time consumed
>(DECps). This seems to me to be rather high, but until I get DECps
>V1.1 I can't tell where this kernel mode time is being spent.
What percentage of the time is the CPU busy?
The DECps figures give a slightly different "picture" from MON MODES.
As an extreme: if you have *NO* user mode processing, then the kernel mode will
and interrupt mode will be very high (percentage-wise).
I don't find 50% to be necessarily extreme. Remember - that kernel
mode time is being spent for some (usually good) reason. You can sometimes
tune for lower kernel mode percentage and get really lousy perceived user
performance.
September, 1993 DECUServe Journal Page 17
(07/14/93 Peach: It "used" to be considered very high)
-------------------------------------------------------
Some of this info may be "old" but it also may be of help:
Excessive Kernel mode time, which in this case means greater than 25%:
- Memory limiation
- Excessive local locking
- High process creation rate.
- Excessive file system activity
- Excessive direct I/O rate
- Excessive use of DECnet
- High image activation rate
Some of my "old" notes have that the Interrupt Stack Time + Kernet Mode
Time should not exceed 40%. But that it is okay if Interrupt < 20 and Kernet
<50 if you done about all the tuning you can do.
Hope this is of some help.
(07/14/93 Holdrege: Our site)
------------------------------
We have a lot of allin1 users on a pair of 7630's with WSEXTENT at
4098. Out of a total of 300%, kernel time is usually around 50%.
(07/21/93 Bonine: Unreliable real-world numbers)
-------------------------------------------------
I just ran some VERY ad hoc (read "next to useless") measurements using MON
MODE. We no longer run DECps, so I can't offer apples-to-apples comparisons.
These were systems running mostly ALL-IN-1 V3 users. I saw numbers which
varied a lot on 1-minutes samples, but almost all of them were well below 50%
kernel mode as a percentage of used CPU during the interval. I'd say we're in
the 30-40% range.
(07/22/93 Scopelliti: Any number helps.. )
-------------------------------------------
Interesting.. we had DEC/CSC dial-in and check out the system - the biggie
they found was a file fragmentation problem. But wait.. we're running
Diskeeper! Looking at the logs.. we _WERE_ running Diskeeper until early
June. At that point it started crashing as soon as invoked. Our disk person
didn't notice...
An upgrade from V5 to V5 cured it.
But... I looked at split and direct I/Os for the period 8-12/1-5 for
the half year (1993) and there really hasn't been much change over the entire
half-year. The ratio of split to direct has remained fairly constant. so.. it
seems unlikely split I/Os are the problem.
We just installed DECps V1.1 - I've got a SYSTEM-PC run for a two hour
period during the heavy morning period, but nothing major sticks out. Lot's of
time spent in SCH$SCHED (no surprise there..), but not much else. Nothing like
10% of time spent in deadlock scans or something like that...
I'd post it, but it's rather lengthy. I'll try to post some key parts.
September, 1993 DECUServe Journal Page 18
A DSNlink article mentions a problem with excessive CPU consumed when
using lots of global buffers (we've got 440 on each major ALL-IN-1 data file),
but it would seem that this would show as high EXECmode time. And we don't see
that anyway...
At any rate.. next weekend we're compressing all our disks and will reduce
the global buffers, so stay tuned.
(07/26/93 Bonine: Tuning global buffers is magic.)
---------------------------------------------------
I would be interested in your experiences with the global buffers. 400
seems high to me, but "sems" is the operative word. Changing it and measuring
seems the best way to find out, but the measurement is very hard. I agree that
I would expect this CPU to be in EXEC mode.
Let us know what you find out.
In Search Of The Perfect Database Package For VMS!
--------------------------------------------------
The following article is an extract of the DECUServe Databases
conference topic 315. The discussion occurred between
July 22, 1993 and August 11, 1993.
By Duncan Brown, Bill Wood, Dale Coy, Simon Maufe, Keith Hare, Richard Norman,
Joe Crum, Mark Shumaker, Seth Grimes, Al Hunt, Carl Friedberg, Patrick Stair
(07/22/93 Brown)
----------------
Lots of people here where I work are convinced that there are tons of good,
cheap, easy to use database packages available for PC's. Paradox, DataPerfect,
and on and on. This seems to be mostly of the form "I heard that..."
So what's available that meets all those requirments, but runs under VMS?
Anything?! We actually *DO* database management as a large part of our
services, but we use an in-house-written database system that is very good at
those large complex tasks. What people want is something that "anyone" can use.
(We'll sidestep the issue of whether "anyone" should be using such a tool if
they know nothing about database design!) While we could certainly write
something in-house again, but make it better suited to small, user-friendly db
applications...that seems like a lot of effort for something that's supposed to
be simple.
Has anyone (DEC, 3rd party, public domain) done this already? Remember,
this has to be "as easy to use as Paradox" (whatever that entails; I haven't
used Paradox myself.) Extra points for something that will run under X-windows
and make people think it's powerful because it's pretty.
(08/22/93 Wood: Consider query and report writing tools against RMS.)
----------------------------------------------------------------------
How about "In search of the perfect single user data storage and reporting
package for VMS"?
September, 1993 DECUServe Journal Page 19
I haven't been real impressed with anything we have considered for general
user data storage and reporting. While Datatrieve is generally too difficult
for many users, it still seems the best choice on the VAX. We have looked at a
couple of screen and windows packages and have not found them anywhere nearly as
easy to use as something like Paradox. (My brain is shot tonight and I can't
remember the names of the packages we reviewed.)
(07/22/93 Brown: It needs multiple users, but it is just store-n-report)
-------------------------------------------------------------------------
If it helps to describe how complex a database tool is needed, let me tell
you how one person prototyped what he was trying to do:
Lotus 1-2-3 WYSIWYG.
Not exactly the king of SQL-based 4GL relational databases, eh?
Like I said, on-screen (and on-paper) looks are at least as important a
factor here as anything else. Fortunately they all realize that the 1-2-3
version doesn't scale up well past about, say, 2 records. But it needs to be
about that simple to throw together.
We have Datatrieve and FMS and such and use them heavily in our "real"
database applications. But there is no way on earth a computer novice is going
to put THAT to use for their own little projects. The classic case of not
wanting to go through the "procedure" of getting "the database guys" to do all
the setup work. It's fine (and entirely NECESSARY) for our big projects, but
total overkill for something tiny like this.
We aren't too heavily into PC's here. Do we need to go buy a bunch of
them and Pathwork them in with the VAXes just so people can do small scale
database work? That seems silly. I thought a VAX could do *ANYTHING*!
(07/23/93 Coy: What will you do with it?)
------------------------------------------
Duncan, what is a "database package"?
There are tons of discussions about several aspects of this on DECUServe.
But I can't decide where to point you.
(07/23/93 Maufe)
----------------
I think the question is a two-part question these days.
o what database engine shall I use?
o what tools shall I use
- breaks down into
o shall I be runnign adhoc stuff
o shall I have an IS staff writing/generating fixed applications
and reports
and, you have to remember what your users have on the desktop.
September, 1993 DECUServe Journal Page 20
I cannot imagine any reason why these days any reasoanble database (>1Gb)
wouldn't be in an SQL based database.
(07/23/93 Brown: Requirements? We needn't state our steeenking requirements!)
-------------------------------------------------------------------------------
> Duncan, what is a "database package"?
Well, now, that *is* the question isn't it. Basically, I have people saying
"we need a database package, but that stuff on the VAX is too hard. We're
computer neophytes, so we need something easy to use. Yes, we know nothing about
computers. So what we need is a Gateway 486/66 with super VGA, Windows 3.1 and
Paradox."
Ahem.
I have been trying to get them familiar with the term "needs assessment" but
am fighting an uphill battle. As near as I can glean out of them, what they
need is ALMOST not worthy of the term "database". If not for just a very few
derived figures (ie math) they could do it all with a text file!
What they basically need is a way to "get stuff into the computer" and then
"get stuff back out again", where "stuff" is just what people used to write down
on a legal pad before computers came into existence.
So it has to be EEEEASY. And at least as important it has to "look neat."
I'm really not making light of that requirement. Hand them Datatrieve and an
RMS file, and it's death. Hand them a mouse, a GUI, and font-filled,
line-bounded, box-shaded PostScript output, and it's a "power tool".
This sort of need comes up all the time. One particular instance now is
going to require this tool to be used by people on the road. Since DEC hasn't
program-announced the Alpha laptops, that means they're definitely going to be
using PCs for the initial interface. But when they all come in off the road and
need to combine their data and share it with others in the company, it sure
would be nice to do that on the VAX! Otherwise we need to build some sort of
PC-LAN just so we can use the PC database package for the combination and
summary tasks.
Have I made this any clearer? I doubt it...
(07/23/93 Coy: Try this)
-------------------------
Well, that's what I thought.
Depending on their background, people tend to have different definitions of
"database". That even includes people who think the most important part of a
database package is how well it does statistical functions.
For VAX packages that have a rather natural interface for data entry, and
other nice things, I really likd RS/1. ForWords will find several threads, but
I might suggest 3RD_PARTY_VMS_SOFTWARE Topic 63 (63.1 is my contribution).
I quite DISagree about SQL for what I understand of _your_ needs. Your
users do not want SQL.
> saying "we need a database package, but that stuff on the VAX is too
> hard. We're computer neophytes, so we need something easy to use.
RS/1 (and some other packages like PV/Wave) attempt to fix that, with very
good results. They focus on getting the numbers IN and getting the reports and
September, 1993 DECUServe Journal Page 21
graphics OUT. [Look at the remarkable absence of the word "database" in the
last sentence] :-)
(07/23/93 Wood: The PC world really does this better.)
-------------------------------------------------------
I know this is not what you want to hear, but I am becoming convinced that
this is an area where the VMS (and for that matter UNIX) packages have just not
kept up with the PC packages. We've wanted to replace DTR for this function for
several years. We've looked at IQ a couple of times, but keep finding it
lacking. However, we do have a requirement that our tool be able to work with
our Rdb databases and that is a tougher requirement. IQ references can be found
in 4GL and 3RD_PARTY_SOFTWARE. It might work better for your environment than
ours.
You didn't define your user community. If you're dealing with engineers/
scientists I have also seen some exceptance of RS/1. If you're talking about
less technical types we've found them totally intimidated by the mathematical
(actually statistical) focus of RS/1.
When I saw Jim Starkey (of DTR fame) at Atlanta Symposium a couple of years
ago he indicated that he was going to start a new company and create a
replacement for DTR for the UNIX and VMS world. I haven't seen any sign of
this, and the last rumor I heard was that he went down the object-oriented
database path instead. I'm quite dissapointed because the VMS environment could
really use a good product in this space.
(07/23/93 Brown: What I expected to hear, I guess.)
----------------------------------------------------
Kind of what I was afraid of.... I do kind of keep up with "things VMS" and
I figured I ought to have heard of something that did this if it existed...
"Technical" would not describe some segmenst of our user community.
They do not want to, for instance, jot down their notes in WordPerfect because
that would be too hard.
If this RS/1 is the RS/1 by BBN, then it would not even begin to suit our
needs. they are the makers of the infamous "Clintrial" software package for the
Pharmaceutical industry. Despite the fact that it ONLY runs on the VAX (using
Oracle) they don't seem to have actually ever SEEN a VAX before... I can not
count the number of ways their installation procedure was broken. In one place,
the error was in a com file generated on the fly! I had to go fix the com file
that wrote the com file, to make it write a com file that would actually
execute... For instance, not using physical device names in all references to
files caused them great problems. I wasn't aware that rooted logicals were such
an exotic concept...
So if they can't even make the INSTALL friendly to a VAX geek, I can't
imagine their PRODUCT could be friendly to a novice. (Of course, I could be
wrong!) That's exactly the kind of senseless stuff I need to avoid at all costs
if I'm to come up with a VAX-based solution for this.
The temptation is strong to WRITE it (there's obviously a market for such a
product, if I was inclined to sell it instead of giving it away!) But I just
don't have the time (in terms of workload, or in terms of when they hope to be
using something.)
September, 1993 DECUServe Journal Page 22
(07/23/93 Coy: It wouldn't hurt to look)
-----------------------------------------
Companies can have different products with different "personalities" and
different degrees of support.
For instance - would you (fairly) judge all Digital software by your
experience with UCX?
So, you might consider looking at RS/1. And maybe one or two of the other
ones I mentioned.
(07/24/93 Hare: Jim Starkey's new product)
-------------------------------------------
Jim Starkey's new product is called Harborview. (or is that the name of the
company? It's too early.) It is currently in development/field test and
scheduled for an initial release this fall.
In terms of building an application with a graphical interface, using a
graphical interface, it looks pretty good. It is designed as an appalication
development tool, not a database design tool, although it does have capabilities
for modifying the database definitions.
It runs on motif/x-windows type devices, with plans for ms-windows in some
future release.
I don't have the information on the company handy, but at some point, I can
dig it out and post it.
Mary Tamir, former DMS counterpart from the TP group is now doing marketing
for Harborview.
(07/25/93 Maufe)
----------------
I didn't mean the usre's would be firing SQL at the database, but whatever
client was chosen should speak SQL for maximum interoperability, now and in the
future.
(07/26/93 Norman: WordPerfect too hard? 8*)) )
-----------------------------------------------
RS/1 is used here for engineering work, but you'd have to write a menu
interface for it to make it usable by novices. Installs can be a pain, but they
have improved.
I would tend to think a hypertext type of data base on the PC would be more
sutiable. I think there are some that take the raw notes and pass it through
filters to create a database. Somebody has to set up the filters and the users
would probably have to stick to a given style which could be forms driven.
At some point you give up functionality for ease of use. I clipped an Ad
that showed a tricyle in front of an F14. Tricyle was labeled "easy to use" and
the F14 was labeled "Fast and Powerful".
Sounds like your bunch won't be happy with anything less than computer
literate stenographers.
September, 1993 DECUServe Journal Page 23
(07/27/93 Crum: Let them have their Paradox...)
------------------------------------------------
I have to agree with the writer of a previous note who said that PC
databases have an advantage when it comes to user-friendliness (I hate that
term), if that's your main objective. If, OTOH, data sharing is the most
important need, VMS databases have an advantage, in my opinion. There may be
some VMS databases I don't know about that are as easy to use as Paradox.
Whatever happened to dBase for VMS? If that product is still a product, it
might be good, but I haven't heard anything about it for a couple of years.
I do development on Oracle, and I also use Powerhouse 4GL, but it sounds like
both of those are too big for this application.
Sounds to me like PCs with Paradox and Pathworks would allow them to do
the simple database stuff, and to share their little databases if they want to.
:-)
(07/27/93 Shumaker: Does Recital still do a dBASE clone?)
----------------------------------------------------------
>> ... Whatever happened to dBase for VMS? ...
Recital Corporation has (or had) what appeared to be a dBASE III Plus clone
for VMS -- at least the user interface looked to be very similar and they
claimed "compatibility" (whatever that means). We didn't investigate it
extensively at the time (about 2 years ago), and I don't know what the current
status is -- or even if they're still in business. Our last contact was in
Danvers MA, at (508) 750-1066.
(07/28/93 Grimes: Harborview info)
-----------------------------------
Harborview is at 508-526-1376 or 7728 (fax). To add some context, James
Starkey designed Datatrieve and the DSRI architecture (according to a brochure I
have) and founded InterBase, which is now sold by Borland. Harborview works on
Unix and VMS against InterBase, Sybase, and Rdb.
(07/29/93 Brown: Anyone DECDecide to use this as a solution?)
--------------------------------------------------------------
I just happened to notice that DEC makes a product called DECDecision that
might just fit the bill. It only runs under DECwindows, but we were already
considering that as a possible factor in ease of use.
Looking under DEC_SOFTWARE doesn't produce a lot of info on this product
(at least just using a title search doesn't.) At least I know now it was late
in arriving, and the butt of many a name-joke!
Anyone used this product as an easy-to-get-going, personal database product?
Anyone used this product period? If the discussion gets involved, we'll move it
over to DEC_SOFTWARE...I just want to know from someone who's usd it if I'm way
off base in considering it as a possible answer to my needs here.
It's a tad pricey, although an ISV discount helps a lot. (SO much so that
I almost think there was a typo in the discount percentage, heh heh heh).
Concurrent-use licenses for it suffer more than most DEC products from the
September, 1993 DECUServe Journal Page 24
"whacked in the head" pricing theory, when compared to capacity licenses on
anything but the very largest platforms. But the capacity license prices might
be swallowable.
I'll also be checking up on the other suggestions made here, of course.
(07/29/93 Brown: Never mind? Or was this JUST a rumor?)
---------------------------------------------------------
I knew it was too good-sounding a product to be true. From DEC_SOFTWARE:
***********************************************************************
Note 577.1 Product Retirement Rumors, DECset etc. 1 of 6
EISNER::HUNT_AL "Alan B. Hunt (Ford Motor Co.)" 5 lines 10-FEB-1993 20:30
-< DECset is safe >-
There have been rumors of everything including DECset, RALLY, DECwrite,
etc being discontinued. None of these are being discontinued. There
is a list out now inside DEC of some products that may be discontinued.
It is very limited. The only two I remember on it are Teamdata and
DECdecision. Most of the rest were PC based versions of products.
(07/30/93 Friedberg: Paradox can talk to RdB?)
-----------------------------------------------
I have not had personal experience with this, but I have been told that
Paradox (DOS version, at least, and several years ago to boot) had an RdB
interface. Any info on that? Any experience?
(08/01/93 Stair: dBASE, no; Oracle, maybe?)
--------------------------------------------
> There may be some VMS databases I don't know about that are as easy to
>use as Paradox. Whatever happened to dBase for VMS? If that product is
>still a product, it might be good, but I haven't heard anything about it
>for a couple of years.
We have dBASE on the VAX and on PCs (both stand-alone and networked via
Pathworkds), and very few users like it on any platform. It is a pain for
multiple users (whatever you networking scheme, whether shared between VAX
users, between PC users, or between PC and VAX users); it imbeds the
path/directory specs in all sorts of files when you create applications (making
it a royal pain to move applications to new directoryies); I don't know anybody
who is a fan of its user interface; and the VMS documentation is not for the
novice. (Do you think, just maybe, that I don't recommend it?).
> I do development on Oracle, and I also use Powerhouse 4GL, but it sounds
>like both of those are too big for this application.
We are seriously considering using Oracle on both the VAX and the PCs,
plus allowing the PC users to use additional things for query and reporting,
plus maybe an additional ad-hoc query and reporting interface on the VAX. What
do you think of Oracle?
September, 1993 DECUServe Journal Page 25
(08/02/93 Crum: But it may be OK for you...)
---------------------------------------------
:-)
Well... I'm not a big fan of Oracle, either. But it works under VMS, and
I've known of folks who used it on PCs a little. In fact, I bought a copy for
my PC once, but couldn't use it because I didn't have enough memory (actually I
bought it just to get the documentation - it was a special price for
developers). And now that I've got enough memory, I wouldn't install it. I'd
much rather use dBase/Clipper on the PC than to load it down with a hog like
Oracle.
I guess most of my complaints about Oracle are a results of some *very*
bad experiences with Oracle products and support a few years ago. Things have
settled down somewhat now, after going through about half a dozen support
specialists and account managers, and upgrading to version 6. Oracle has a very
good marketing strategy. They sell you the base product and lots of tools to
allow you to use it, then they sell you training. Then the next year, they
completely change the database and/or tools (adding functionality, of course),
so you must forget everything you knew about the previous version and go back to
school (becoming a novice again in the process) to learn all about the latest
and greatest changes. Last year I took the SQL*Forms 3.0 advanced developers
course. Now we are seeing ads for the version 4.0 course, which is, according
to some Oracle folks, completely different than 3.0. Of course, 3.0 was
completely changed from 2.3. They keep changing all of this stuff while
ignoring requests for things like RMS file access under VMS (which I understand
may now be available as an optional product), which to me should be a part of
any 4GL tools like SQL*Forms under VMS. I admit I'm a bigot when it comes to
databases - I just haven't seen a compelling argument for using them under VMS.
Or maybe I should say I don't think they are good enough yet to justify the
money you pay for something like Oracle, when RMS files will do basically the
same job for free. And I have very rarely needed to "re-index" an RMS file
because the index was corrupted, or to shut down all of our applications to do
an 8-hour "export and import" to fix a corrupted RMS file. We run a 24*7 shop,
and can't afford to put our critical applications in a database system which
requires periodic downtime (no matter what the Oracle reps say). *flame off*
Arent't you glad you asked? :^}
(08/10/93 Brown: It's only money!)
-----------------------------------
I got a 60-day demo of DECdecision from the E-store. So far it looks
pretty slick, and might be just the kind of thing we're looking for. Of course,
it ain't cheap...ad in typical DEC fashion, you may need to buy a few more
things to make it do what you thought it was going to...
See my note on DECdecision in DEC_SOFTWARE for the ongoing saga.
(08/11/93 Brown: One way to look at it!)
-----------------------------------------
> Harborview is at 508-526-1376 or 7728 (fax). To add some context,
> James Starkey designed Datatrieve and the DSRI architecture (according
Finally received the Harborview info today. I haven't read it in detail
(though there ain't much there!), but I did note that James Starkey considers
HarborView to be "Touchy-feely-trieve". Ar ar ar.
September, 1993 DECUServe Journal Page 26
Macintosh network backups?
--------------------------
The following article is an extract of the DECUServe
Personal_Computing conference topic 496. This discussion
occurred between August 9, 1993 and August 11, 1993.
By Clay Denton, Larry Kilgallen, Bob Koskovich, Saul Tannenbaum, Joe Bates,
Mike Durkin
(08/09/93 Denton)
-----------------
Is anyone familiar with a product to automate backing up networked Macintosh
systems to a VAX based tape drive or an AppleShare server (which could be a VAX
that could then place data on tape).
Or, if anyone has home grown procedures I'd be inteested in what you set up.
Thanks for any ideas...
(08/09/93 Kilgallen: Retrospective Remote)
-------------------------------------------
I am under the impression that Retrospective Remote by Dantz is the
definitive product in this space.
(08/09/93 Denton: Contact info?)
---------------------------------
Do you have contact information?
(08/09/93 Koskovich: Retro. Definitely.)
------------------------------------------
Retrospect is a snazzy product. I use it every day on my Mac.
It'll back up to most SCSI-connected tapes local to the Mac, as well as any
mountable *anything* (like AppleShare volumes). I'm pretty certain that you
won't be able to back up to a VAX-connected tape drive.
They do a good job of tech support.
Published by:
Dantz Development
4 Orinda Way, Bldg C
Orinda, CA 94563-9919
510-253-3000 (main)
510-253-3050 (tech)
510-253-9099 (fax)
September, 1993 DECUServe Journal Page 27
(08/10/93 Tannenbaum)
---------------------
Another vote for Retrospect/Restrospect Remote.
It was good in Version 1.
Version 2 makes it even better.
(08/10/93 Bates)
----------------
But with a little bit of magic with ResEdit Retrospect will backup to a
TK50/TLZ04/TLZ06 SCSI tape drive connected to a Mac.
Also if you have PATHWORKS for Mac and DECnet/Mac installed on your
Macintosh, you can use DECnet to backup Mac hard drives to a VAX based tape
drive. It's a bit kludgy, and not necessarily screaming fast, but can be done,
is documented in the PW4M manuals and PW4M provides an example DCL command
procedure to do it with installation of the PW4M server.
(08/11/93 Durkin: TK50?)
-------------------------
> But with a little bit of magic with ResEdit Retrospect will
> backup to a TK50/TLZ04/TLZ06 SCSI tape drive connected to a Mac.
I was attempting this a while ago and thought the limiting factor was the
absence of an appropriate driver for the TK50 ? Back in thread 354 here, it
looked like it was possible, however the Dantz support staff indicated the TK50
was not a supported unit. I'd be most interested in obtaining a copy of your
ResEdit hacked version or a detailed explanation of your ResEdit hack. :-)
The Art and Science of Telemarketer Baiting
-------------------------------------------
This article is an extract of the DECUServe Hobbies_and_Interests
conference topic 109. This conversation occurred between
June 18, 1993 and July 30, 1993.
By Sharon Linnea Johnson, Pat Scopelliti, Jamie Hanrahan, Linwood Ferguson,
Ta Fuh Chiam, Gary Rice, Chris Simon, Malcolm Dunnett, Bart Lederman,
Bret Wortman, Jane Furze, Charlie Wrenn, Jean-Francois Mezei, Larry Kilgallen,
Matt Holdrege, Eric Dungan, Rob Brooks, Dan Wing, Bob Koskovich, Ray Whitmer,
Rick Carter, Bruce Bowler, Al Lilianstrom, Mark Shumaker, Terry Kennedy,
Richard Norman, John Gorentz, John Lee Ellis, Gus Altobello, Jamie Hanrahan,
Bob Hassinger, Michael Baydoun, Kevin Hoot, Al Hull, Charlie Luce Jr.,
Glen Zorn, Bill Wood, Kevin Angley, Herb Mattord, Brent Hutto, Roger Bruner,
Craig Snoeyenbos, Larry Horn, Curt Snyder, Jack DelBalso,
(06/18/93 Johnson)
------------------
I guess telemarketer baiting is a kind of hobby. I think that since these
people are invading my privacy I am entitled to try a few creative things to
September, 1993 DECUServe Journal Page 28
discourage them from calling me again. However, abuse is against my ground
rules.
We got two of those telemarketing calls trying to get you to sign up for a
sales pitch for some kind of timeshare condo. (Standard spiel is to tell you
have won one of five prizes, then collect some demographic info on you, such as
age and income so that they know it's worth trying to rip you off, and then sign
you up for an appointment to listen to some sales pitch.)
The first one was for me. I got through the questions up to the point where
they try to make the appointment. At that point I set the phone down on the
table and went out to work on my garden for about an hour or two. For a while
you could still hear the background noise on the line. Later the line just hung
there, but it did not hang up. After about two hours, we had to hit the
switchhook twice hard to hang it up from our end. From this experience I have
concluded that they have had this one before and it probably is not too
effective. In fact, it was obvious that the same outfit called back our number
about a week later...
She asked for my husband by his first name when she called. Since I could
hear the heavy phoneroom background noise, I put him right on. He was just
waiting to try his latest idea for baiting telemarketers. As soon as she listed
the five prizes (one of which he won), he started on a routine of "do I get to
pick which prize?", and "I want the $25000 cash". He put on a very naive and
enthusiastic act and kept digressing her back to these two main points. He
pretended he just didn't understand why he needed a 90 minute appointment to
just go over there and have the money handed to him. He managed to keep at this
for about 10 minutes. She had to put him on hold and talk to a supervisor
twice! Finally she told him that her shift was over and she had to go home and
hung up on him.
Recently we also got one of those calls from a large metropolitan newspaper
well known for trying to sell you a subscription over the phone (even though you
subscribe already). We got to use the line someone gave us that we had been
saving up for this. (Really we don't subscribe; the bird does.)
Telemarketer: I'm from the <newspaper name here> and
Respondent: (interrupt with great feeling) I DON'T DO INTERVIEWS! (hang up
immediately)
I am interested to know what other folks have tried as methods to discourage
this sort of thing. (Language which violates canons does not count.) I think
my husband's latest one is based on the idea that if you waste their time by
talking and talking and never buying a thing that it reduces their profits.
P.S. My previous method wasted very little of their time because it
consisted of simply saying "No, thank you." and hanging up. (Minnesotan for
"I'm not interested".) A couple weeks ago I got a call from a guy who just
didn't understand Minnesotan. Naturally, I was so startled, I'm afraid I
sounded a bit surly to him.
Telemarketer: I'm <name> from <company>.
Sharon: No, thank you.
September, 1993 DECUServe Journal Page 29
Telemarketer: (somewhat surly) No thank you what?
Sharon: (sternly) No thank you, I'm NOT INTERESTED. (hang up)
(06/18/93 Scopelliti: Your local fish wrapper cold call)
---------------------------------------------------------
Ah yes.. some fun!
I used to get called about once every three months and immediately told:
"Hi, I work at the Elmira Star Gazette and we're checking to make sure
your paper arrived in good condition and on time today."
After telling them I don't get the paper, they usually launch into the
standard spiel "Oh, would you like to get it for only ..."
Last time I finally figured it out - I answered "Yes, it was just fine!"
Now, they _KNOW_ I don't get the paper. The next thing I hear is the sound of
gears grinding followed by a confused voice saying "Oh, OK.. thank you"
Haven't heard from them since!
(06/18/93 Ferguson: Just the FAX)
----------------------------------
Different but Related: I didn't find most of Bill Hancock's book funny (the
one they gave away at the trade show), but for those who send you marketing FAX
with a return 800 FAX number I loved his response:
Write a long note. Start it FAXing, and as it goes through tape the two
ends together making a loop. Leave for the day.
Actually this might be worth it even if you pay for the call.
(06/18/93 Chiam)
----------------
Ha I love this one!
My technique is to stall them for as long as I can, if I have got nothing
better to do at the time.
Here are some of the things my friends and I did:
1. Say "Hold on, my water is boiling" then hand the phone to a 3 year-old
and let them talk. :-)
2. One of my friends once said "Gee, you are the third person who called me
today wanting me to buy something. What? Do you think my father prints
money or something?" The telemarketer said "sorry" and hung up.
3. Another friend said "No, I can't talk right now, I am making <potential
canon violation phrase here> with my wife."
4. When I have got nothing better to do, I will just stall along while I
watch TV. I keep saying "wait, can you repeat that please?" or "Wait,
can you slow down, you are going too fast and I don't understand." After
a while, I will then say "No, I don't think I am interested."
September, 1993 DECUServe Journal Page 30
5. I will say "Oh, <newspaper name>, no I don't subscribe to it and I don't
want it." If asked why, I will say "My next door neighbour gives her
paper to me to read first so that I can tell her what is new and I love
doing that! If I subscribe to it, I have no reason to go see and talk
to my lovely neighbour."
6. I will sometimes pretend that I don't understand and speak English very
well at all. :-)
The list can go on.
(06/18/93 Rice)
---------------
> Telemarketer: I'm from the <newspaper name here> and
Come on Sharon, 'fess up. I'd bet that the "<newspaper name here>" was
the LA Times. I've gotten 2 or 3 calls from them every year for as long
as I can remember.
Well, in the case of the Times, I have a standard speech:
"When you guys stop running ads at the local Cimena (and making me PAY to watch
your drivel), I will consider subscribing. Until that happens, don't bother
calling back... "<click>"
(06/18/93 Simon: Newspapers are a pain)
----------------------------------------
Back when Dallas still had two newspapers, both of them called very
regularly to ask if we wanted to subscribe (we already had subscriptions to both
of them). I tried everything I could think of to get them to leave me out of
their little circulation war. The thing that finally worked: I told both of
them that if they _ever_ called me again to see if I wanted a subscription, I'd
cancel immediately.
(06/18/93 Hanrahan)
-------------------
John Higdon, a frequent contributor to comp.dcom.telecom, can top almost
everybody's stories along this long, with his tales of harrassment by the San
Jose Mercury News. They used to call his phone lines regularly -- every one,
including the personal and business lines, the TWO fax modems, and several lines
used for UUCP and PPP. They'd hit each one in turn. John told them several
times to not call again. He told them that if they called again he'd see about
suing for harrassment (having told them previously that he DID consider it
harrassment, and please stop). He finally had a "lawyer letter" sent to them.
This stopped the calls for about a year... and then whoever was in charge of
telemarketing got replaced, or forgot about John's "requests", or something, and
the calls started again...
The LA Times used to bother us regularly when we had a Sunday-only
subscription. Now, you need to know that Sun-only subs are a BIG pain for the
September, 1993 DECUServe Journal Page 31
newspaper. But they do make money and the newspaper keeps hoping that you'll
convert to a full sub... and they keep bugging to do so.
"We've got a special deal -- you can have a month of the full delivery at the
Sunday-only rate. Then you can see if you'd like to continue..."
"Look, if I WANTED a full subscription I'd HAVE one. I simply don't have time
to read that much, and I don't want the paper delivered just to be thrown out."
"But ... we're doing this with ALL of our Sunday-only subscribers."
"You mean I simply can't continue getting the Sunday-only?"
"No."
"Well, I'll just have to cancel the whole subscription then."
"Uh... I'll see what I can do..."
This went on at least twice a year.
Having said all that, it IS always worth remembering that the poor person
who's calling you is not the person who DECIDED to call you, and furthermore
would gladly take almost any other job if possible. The advice to try to get
the person's supervisor on the phone is good, but even there you're likely
dealing with a supervisor at a telemarketing firm, not the person who's the real
problem...
(06/18/93 Dunnett: Time to write a new speech? :-) )
-----------------------------------------------------
>the LA Times. I've gotten 2 or 3 calls from them every year for as long
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>to watch your drivel), I will consider subscribing. Until that happens,
>don't bother calling back... "<click>"
^^^^^^^^^^^^^^^^^^^^^^^^^
(06/18/93 Lederman: You CAN stop telemarketing with this simple method.)
-------------------------------------------------------------------------
Odd, the "L.A. Times" never bothered me with telephone calls when I had
a Sunday-Only subscription.
Of course, I was living in Brooklyn at the time, and got the paper by
mail...
(06/18/93 Wortman: Tell the truth!)
-----------------------------------
The ones who seem to call me most often are window companies. Odd, since I
used to work for one....
However, I've had more fun just walking the spiel, getting all the info
out of them I can possibly get, then actually giving them my address.
##### W. 64th Terrace, Apartment 511.
Works great with shinglers, new-room-addition contractors, and so forth.
September, 1993 DECUServe Journal Page 32
(06/18/93 Furze: One of my favorite topics)
--------------------------------------------
If I have time, I try to stall them for awhile with what ever my latest
favorite technigue may be. For example, if they just want to ask me "a few
questions" it usually runs like:
Caller: What is your name? Do you own your home or rent? etc.
Response: That's a very personaly question. I'm not in the
habit of discussing my personal matters over the phone
with total strangers. <pause> Wait a minute.
Is this an obscene phone call? You people really are
disgusting! You know what you're doing is illegal.
I must inform you that this call is now being recorded. ...
Caller: Are you married?
Response: Why? Are you asking me out? I never date strangers.
Your know, you really should not make dates with people
you've never met. It can be quite dangerous.
If the caller happens to be female, canons prohibit
my describing my response.
Caller: What is your sex?
Response: Oh, I abstain. You know, with all the strange stuff
on the streets, that's the only really safe thing to do.
Caller: Are you male or female?
Response: Does it matter? This sounds like a sex discrimination
issue to me. How long has your company been openly
practicing sex discrimination?
If I'm busy I just point out that this is a private line & I paid the bill
for it last month. When their company starts paying my phone bill I am willing
to discuss with them whether I will permit them to use my phone line to conduct
their company business. Please have someone in their company, with the proper
authority, contact me to make the necessary arrangements.
I have a friend who has been a widow for 20+ years but has left the phone in
her husband's name. When someone calls for Paul she plays with them for
awhile but tries to be nice since she has worked for minimum wage in the past,
herself. Once in awhile the caller gets 'assertive' and insist on speaking to
Paul. She considers these people fair game for anything she wants to dish out.
On a couple of occasions, the caller has been so foolish as to claim "Paul told
me to call him." Her response to that one was to get really concerned and sob
"Oh really? Well would you please do me a favor? Next time you talk with him,
please tell him I want to speak with him as soon as possible. You see, the last
I knew, his ashes were still buried under a rose bush in my back yard." That
has always succeeded in ending the conversation.
(06/18/93 Wortman: Other phone fun....)
----------------------------------------
I just love that "To whom am I speaking?" question.
"The person you called, silly!"
September, 1993 DECUServe Journal Page 33
For a while there, I used to answer the phone with (pick one from the list)
"Thank you for calling Pizza Hut Delivery, may I have you telephone number?"
"Thank you for choosing AT&T."
"Hi, is Bob there? No? Sorry, I must have the wrong number." [hangup]
(06/18/93 Wrenn: BILL THEM!)
-----------------------------
I have found this quite effective. Play their own game.
Me: Sure you can come to my house and demo <product>. But first let me get
some information. I'm a consultant and charge $XX per hour, with a minimum
charge of $XX. Let me get the address so I can send you an invoice for
this phone call also - I charge $XX for phone consultations.
Them: (almost every time): <click>
(06/18/93 Mezei: Call waiting is the best answer to telemarketers)
-------------------------------------------------------------------
The one that works the best for me, and does not sound overly nasty to the
poor person who must call is:
You sound really interested. Then, when they start to want real information
from you:
"Can I put you on hold for a minute? I got another call waiting, I'll
be right back"
You are telling them they are important because you will be back and that
other call isn't as important as theirs. Yet, you do click your phone call some
number and then just hangup. The poor person will wait for a while, and when the
line finally drops he will realise he has been had....
I even saw a gadget for sale on TV one night that would produce the beep on
your end and the "silence for half a second" at the other end to really simulate
the call waiting feature.
(06/18/93 Kilgallen: Massachusetts is Automated)
-------------------------------------------------
You are to be congratulated for starting this topic.
I guess the rest of the country is not very well automated, because
all I ever get are _automated_ solicitations, which record a response for later
scanning by a human. I have thought of, but never used, "I realize the person
listening to this is not the owner of the company, and may not be adequately
paid for taking part in bothering people. I will pay you $x (adjust for local
economy) to not report this response to management but instead call me from
elsewhere and give me the home phone number of the owner of the business."
From there on, I would not even need DECtalk to harrass back -- a modem
would do.
September, 1993 DECUServe Journal Page 34
But no, I am too cheap, or lazy, so I generally just hang the headset
on the television next to the speaker to use up their tape. Someday I will get
a gadget so I can just transfer the call the the television.
> noise on the line. Later the line just hung there, but it did not
> hang up. After about two hours, we had to hit the switchhook twice
Typically, if you remain off hook when your caller hangs up, the phone
company equipment puts you in a special wait state which takes longer (60
seconds or so) to get back to normal after you go on hook.
Some old step-by-step exchanges allowed a caller to "pin" the phone of a
called person, but I have never heard of a situation where the called person
could actually "pin" the phone of the caller. So it is just you against human
endurance, unless the calling company has arranged things so their employees can
never be the first to hang up. Of course, there is always the shift change :-)
(06/18/93 Lederman: The Source.)
---------------------------------
There is a very good book which is the definitive source on this and related
subjects. It's called "The Compleat Practical Joker", by H. Allen Smith, ISBN
0-688-03705-4. This is the revised version, the original came out in 1953.
Several chapters have to do with the telephone. Of course, this was written
before telemarketers, but the chapters on the telephone also explain double-talk
and the mumbles (they're different) and how they can be used, and some of the
treatments for wrong numbers should work well on telemarketers.
The chapters on Hollywood, New York (did you know Ed Sullivan was a
practical joker?), and indeed the entire book, are quite good as well.
(06/18/93 Scopelliti: Instant filter!)
---------------------------------------
I've got a second line at home for dialup use. No one I know ever calls
that number.
So, when it rings, I know it's gonna be fun!
e.g.
"Hello, Corning Morgue"
(06/19/93 Wortman: Real World Experience -- Just last night!)
--------------------------------------------------------------
Talk about synchronicity. Last night I kept an AT&T operator on the phone
for half an hour before asking if he could change both my fiancee and I over to
AT&T. He said he'd have to talk to her to get her approval.
Unfortunately, I told him, she was at work at the moment but he could
probably reach her there. "What does she do?" he asked, already having
ascertained that I was a computer geek twenty minutes ago during one of many
sidetracks.
"She's a directory assistance operator for Sprint."
September, 1993 DECUServe Journal Page 35
It didn't take long for him to hang up after that. Oddly enough, he got a
little snippy with me. Can't imagine why....
(And I ain't even engaged!)
(06/19/93 Mezei: I am sorry, there is no service at the number you have dialed)
--------------------------------------------------------------------------------
Once I picked up the phone said Hello? And then they made it clear it was
a telemarketing call. I immediatly said "I am sorry, there is no service at the
number you have dialed, please check your number and try again later". They
hung up.
Interestingly enough, I live an an English portion of Montreal, and I get
pestered by the English newspapers. Never by the French newspapers. So, when I
get pestered by English newspapers, suddently, I no longer speak nor READ
English. Even though the person will switch to French and try to sell me the
English newspaper I insist that I can't read English and tell them they are
wasting their and MY time.
Once, I spent about half an hour on the phone with a guy (again for an
English newspaper) and we had interesting arguments. He really wanted me to try
his FREE 3 week offer and that I had nothing to loose. I kept telling him I was
not interested. Finally, I told him:
"OK, just so that you get your commisssion, I'll subscribe to your offer
with one condition: you must give me the telehone number I can call to cancel
the subscription immediatly". He did. When the call was over, I called and
complained to the subscriptions people that the harrassemenet was getting out of
hand. They agreed to cancel my free 3 week offer. I ended up getting it
delivered for about a week.
(06/20/93 Holdrege: My little trick)
-------------------------------------
I used to work for the Chicago Tribune and I still got calls to subscribe.
When I told the caller that I work for the Tribune and I get the paper free, she
chuckled and said goodbye.
Now, when the LA times calls me I just say that I work for the Times and
I get the paper free. They always let me off when they hear that.
(06/21/93 Dungan: TM Retort from "SEINFELD")
---------------------------------------------
Funniest retort to a tele-marketer that I have heard/seen was on an episode
of Seinfield:
Phone rings, Jerry answers
TM: Hello, this is <JUDY> from <X-Long Distance Company> & I would like to
talk to you about switching to our service.
Jerry: I'd like to talk, but I have some guests right now. Can you give me
your home phone number and I'll call you later after my guests leave.
September, 1993 DECUServe Journal Page 36
TM: Oh, I can't do that, we are not allowed/it would be disturbing us to
call us at home
Jerry: Now you know how I feel. CLICK
I've tried this and it works.
Encrypting a disk at the physical level
---------------------------------------
The following article is an extract of the DECUServe Security
conference topic 257. The discussion occurred between
July 19, 1993 and July 26, 1993.
By Jean-Francois Mezei, Dan Wing, John Burnet, Rob Brown, Michael Baydoun,
Gus Altobello, Scott Dawley, Rick Carter
(07/19/93 Mezei)
----------------
How realistic would the following scenario be:
Something triggers a "safagard my disks please" and the following happens:
-All non critical processes on the system are stopped.
-A program then proceeds to encrypt all target disks
at the physical level (eg: with PHYIO and encrypts every
physical block of the disk).
When the situation has cleared, the disks would then be decrypted in the
same manner.
1- How vulnerable would the encryption be considering that quite a few
blocks would be empty (and contain nulls) ?
2- Is this feasable anyway ?
The reason I ask is that I do not normally require encryption. However,
in the ~unlikely~ event of a break-in :-( , I would like to zap my disks before
anything could be done to them, but in such a way that I could recover them
later.
I realise that this is not foolproof, but it does provide a safegard that
"looks" good. Something similar to a self destruct mechanism on those science
fiction ships : destroy your data to prevent them from stealing it.
(07/19/93 Wing: Just erase the data?)
--------------------------------------
INIT/ERASE would give you that functionality. And I seriously doubt you'd
have a physical break-in that you would have enough time to hit the 'secure
the data' button. If they are that close to you, they could hit the RESET
buttons on your VAXes!
September, 1993 DECUServe Journal Page 37
If they are coming in over the network, unplugging your Ethernet cable is
a good short-term solution to preventing your data from getting stolen.
(07/20/93 Mezei: Physical access means logical access to the data)
-------------------------------------------------------------------
My application for disk encryption would be triggered by an alarm system
and would offer philosophical protection against an intrusion while in "lights
out mode" (eg: while I am out cycling or doing other stuff not related to
computing ).
The goal is to offer a certain level of apparent logical security in the
event of a physical security failure (eg: break-in).
Remember that gaining physical access to a VAX gives you access to all of
its data.
Even if not all the disk is encrypted by the time the criminals get to it,
it is bound to be damaged enough to make it difficult to access the info easily.
I assume that INDEXF.SYS normally sits at the beginning of the disk and would
be one fo the first files to be zapped by an physical block by physical block
encryption of the whole disk.
I must also be concerned with Personal Computers. How do you really secure
the information in a PC to make it inaccessible when physical access to a
machine is gained? From what I have read, opening the PC and removing its
battery is usually enough to remove all password protections that prevent
access to the PC.
(07/20/93 Burnet: Start protecting your data *before* it's too late, not after)
--------------------------------------------------------------------------------
> I assume that INDEXF.SYS normally sits at the beginning of the disk and
>would be one fo the first files to be zapped by an physical block by physical
>block encryption of the whole disk.
The default behavior is to put the index file in the middle of the disk.
You can override that with INIT/INDEX=BEGINNING.
It seems to me that trying to encrypt or erase a mounted ODS-2 volume while
the system is running is asking for trouble. What do you do about processes
with open files on that volume -- especially for the system disk (which
can't be dismounted)?
I think you would be better off using something like Glenn Everhart's
virtual-disk driver (it's on many DECUS tapes and CDs) to keep all of your
"secret" files encrypted at all times. Then, if someone breaks into your
building and takes control of your VAX, you can let him take whatever he
can get his hands on (virtually speaking) and you won't have to worry about
it too much. If all of your own stuff is encrypted (project files, clients'
files, or whatever), then an intruder would only be able to steal copies of
VMS and any other software that's installed.
(07/20/93 Burnet: Just to put this in perspective... )
-------------------------------------------------------
I hope this doesn't sound too rude, but an analogy comes to mind:
Let's say that my wife and I have many tens of thousands of dollars worth
September, 1993 DECUServe Journal Page 38
of jewelry, negotiable bonds, and cash in our house, and that we keep all this
stuff in easy-to-find places like unlocked desk drawers.
We're planning to go on vacation for a few weeks, and I would like to set
up a robot that's connected to the house's security system so that if someone
breaks in, the robot will immediately collect all of the valuables and lock
them up in a safe. How can I do this, or where can I find such a robot?
(07/20/93 Mezei: Maybe I should just look at INIT)
---------------------------------------------------
Vacations are one thing. But being away for an afternoon is another. I
resist using permanent encryption because of the raw horsepower that my all
mighty MicroVAX 2 just doesn't have. Of course, it would be the ideal
situation, but woudl require additional management.
But I would like some sort of method to make data useless if an alarm is
triggered.
I want "protection" that is more of less automatic. When you go on vacation,
you know you are at risk and you take precautions such as hiding jewelry etc.
But when you are away for an afternoon or a couple of days, you generally don't
want to hide everything. In the case of data where my vax is up 24 hours a day
7 days a week even while away on vacation, I just want a little pillow of
comfort knowing that is something happens while I am away, I can tell my clients
that their data wasn't compromised. It isn't data that is worth encrypting
constantly, but I wouldn't want to see it stolen and risk that it be used
elsewhere.
I think that the INIT command may end up being the most useful in the end.
I just have to make sure I take proper backups.
(07/20/93 Brown)
----------------
Instead of trying to encrypt the whole thing, why not just damage the home
block so that the disk cannot be mounted. It occurs to me that the easiest
thing is to just change the pointer to the the index file bitmap.
Create your recovery program as a bootable system on your TK50.
(07/21/93 Baydoun)
------------------
Don't forget to lockup your backup tapes.
(07/21/93 Altobello: ...better be *good* backups...)
-----------------------------------------------------
Don't forget to keep your backup tapes *somewhere else*.
(07/23/93 Dawley: Good looks, or real security?)
-------------------------------------------------
> I realise that this is not foolproof, but it does provide a safegard
>that "looks" good. Something similar to a self destruct mechanism on
September, 1993 DECUServe Journal Page 39
Here's just the problem...it LOOKS good, but it won't really DO any
good. My reasons:
1) You are hoping that your blaster device will have enough
time after the physical breaking to encrypt everything.
Not something _I_ would like to assume if my data is that
critical.
2) If someone is really really after your client's data,
they'll do the breakin right to begin with, and your
blaster will probably never even get the chance to do it's
work. (Example: what happens if your system is deprived
of power before the physical breakin?)
3) You are attempting to do through technology what should best
be accomplished through good physical security. Put your
system with the 'secure' data on it behind several more
locked doors. Then MAYBE your blaster will have time
to encrypt your whole disk.
4) Hire a security guard...or buy a nasty junkyard dog
and throw him some meat now and then for training.
Don't do things just because they LOOK good...make sure they will really
do what you hope they will do.
(07/26/93 Carter: Why not just spin down the disks?)
-----------------------------------------------------
September, 1993 DECUServe Journal Page 40
Technical Information
---------------------
Publication Information
-----------------------
Topic threads in the DECUServe VAXNotes conferences are selected for
publication on the basis of strong technical content and/or interest to a wide
audience. They are submitted to the editor by a team of Contributing Editors
who are DECUServe Moderators, Executive Committee members and other volunteers.
Articles in the DECUServe Journal are downloaded to a 286 PC and formatted
using a standard text editor.
Editorial Content Disclaimer
----------------------------
Opinions and information presented here belong to each individual author.
No inferences should be made that the authors are expressing the policies of
their employers unless specifically stated. Content has been deemed acceptable
by the Moderators of DECUServe according to the DECUServe Canons of Conduct.
How can I use DECUServe?
------------------------
DECUServe is available 24 hours a day, 7 days a week, except for the last
Thursday of each month at 5pm Eastern time for backups. Membership is by
individual subscription only, and subscriptions are sold on a yearly basis
currently for $75.00. Subscription forms are always included in DECUS New
Member packets and Symposia registration packets. Six-month subscription
forms are often distributed at symposia as well.
Foreign DECUS chapter members are invited to subscribe as well, although
they must do so through their chapter office. Forms and price details can be
obtained from the chapter office.
Also subscription forms can be downloaded directly from DECUServe by dialing
1-800-521-8950 with the username INFORMATION. Set your modem to 2400 or 9600,
1 stop bit, 8 bits, no parity. Internet access is also available.
How can I get The DECUServe Journal?
------------------------------------
The Journal is now available by Internet Mailing list. Simply send email
to "mail...@decus.org" with the words, "subscribe decuserve-journal" as the
body of the message.
It is also posted monthly on the vmsnet.decus.journal and comp.org.decus
newsgroups, and on CompuServe.