D3 Spooler Wonder
388 views
Skip to first unread message
Peter McMurray
Dec 5, 2015, 5:34:06 PM12/5/15
to Pick and MultiValue Databases
I have just spent an interesting time trawling through checking printers and even reading the D3 manual :-)
My prints suddenly stopped coming out. Listpeqs showed them , listptr showed them, I rebooted still no joy.
Finally I did what I should have done in the first place :-( SP-STATUS
:sp-status
The spooler is inactive.
Needs to start printers.
Needs to start printers.
Printer # 0 is serial, inactive, and on line.
The printer is running on line 31.
Assigned output queues: 0 .
The number of inter-job pages to eject is 0.
The printer is running on line 31.
Assigned output queues: 0 .
The number of inter-job pages to eject is 0.
Printer # 10 is serial, inactive, and on line.
The printer is running on line 10.
Assigned output queues: 10 .
The number of inter-job pages to eject is 0.
:startspooler
[1139] The spooler is started.
:
[1139] The spooler is started.
:
Now there is a mystery printers clearly started but the spooler saying they are not. I believe that is the first time in 38 years that I have used startspooler at TCL.
It worked so I wont knock it. Any explanations?
Peter McMurray
Dec 6, 2015, 9:34:13 PM12/6/15
to Pick and MultiValue Databases
Following a short discussion with MBS support I found line 0 was still on somehow, detail as follows, Ross may be interested in the para about missing frames taken from the D3 manual
I have fixed the problem. I noticed both Startspooler (c and Startspooler (I in the errors file ( I have been shutting down several times to solve the issue) and I found that line 0 was logged on so something went astray.
Anyway I did a quick VME restore and did NT_Upgrade again and all is fine.
In the process I believe I have solved the issue of missing frames that has been discussed on the user group.
Startspooler (i
Starts the spooler and initializes control data with new overflow frames. The disk space used previously by the spooler is lost until the next file restore (approximately 5-9 frames). This is the last resort in resetting the spooler.
Ross Ferris
Dec 7, 2015, 5:57:05 PM12/7/15
to Pick and MultiValue Databases
Is
STARTSPOOLER (I
part of colstart? Don't think this is a "solution", and indeed the exercise probably does more to reconfirm the presence of Gremlins!!
Tony Gravagno
Dec 7, 2015, 8:32:54 PM12/7/15
to Pick and MultiValue Databases
Nice try, no cigar.
When people talk about missing frames they don't mean <10. They're talking about a 250MB VME getting eaten up every week. So much and so often they had to script the operation because it was the same all the time and tedious - and if you need to do this for a bunch of end-users it's downright painful. No one is doing that many spooler restarts where they're going to lose that many frames on a frequent basis. Think about it - you've never used that function in your decades of experience, and most others don't either.
I have no idea what the state of this issue is anymore. TL/RS said they believe they fixed it, or perhaps the statement was more like they believe they've alleviated most of the issues, so a current release might not exhibit this issue at all.
RE: "This is the last resort in resetting the spooler."
Every time I see that I have a laugh. In EPick, aka AP/REF, Jon Sisk wrote "This is the last resort in spooler warfare", and there were other gems in there like that. I regret that on later review I had some of those rephrased to exhibit a bit more professionalism. But in my head I still think "warfare". LOL
T
When people talk about missing frames they don't mean <10. They're talking about a 250MB VME getting eaten up every week. So much and so often they had to script the operation because it was the same all the time and tedious - and if you need to do this for a bunch of end-users it's downright painful. No one is doing that many spooler restarts where they're going to lose that many frames on a frequent basis. Think about it - you've never used that function in your decades of experience, and most others don't either.
I have no idea what the state of this issue is anymore. TL/RS said they believe they fixed it, or perhaps the statement was more like they believe they've alleviated most of the issues, so a current release might not exhibit this issue at all.
RE: "This is the last resort in resetting the spooler."
Every time I see that I have a laugh. In EPick, aka AP/REF, Jon Sisk wrote "This is the last resort in spooler warfare", and there were other gems in there like that. I regret that on later review I had some of those rephrased to exhibit a bit more professionalism. But in my head I still think "warfare". LOL
T
Peter McMurray
Dec 8, 2015, 3:27:39 PM12/8/15
to Pick and MultiValue Databases
It seems that there is a major error history that I have never met. If people were losing 250Mb at any time that would be enough to dump the product. However Ross talks about minor leakage and the test he suggests does not in my experience lead to anything more than the few frames I found.
The major error must have been in the far distant past and long since fixed. I do remember that the spooler did not scale well 20 years ago causing a problem in the States for a chain of health clubs if I remember. But D3 on Windows in my experience goes well and this frame problem has never been discussed at any of the many meetings I have been to.
Regarding my problem, it is quite likely that I made a restart error when playing around looking for the missing frame fault :-(
Tony Gravagno
Dec 8, 2015, 8:30:11 PM12/8/15
to Pick and MultiValue Databases
When companies have serious issues, they usually don't take them to forums. They contact their vendor and do what they can to collaborate toward a resolution.
Similarly, when companies leave a platform they usually don't advertise the fact, burning bridges and embarrassing themselves with a public rant on their way out the door. They sometimes don't even tell their vendor, who may be made aware of a migration only when the Support renewal is not processed.
The fact that you did not know about this should not trouble you. Nor should you feel inclined to dismiss the severity, with the personal good fortune of not having experienced this yourself. Several people in this group have experienced the described problem - it's up to each individual to decide what they publish. And yes, many sites Have migrated, either from D3 Windows to Linux, or (dumped the product and) moved completely away from D3. This was not in the distant past, and was only fixed within the last few years ... and as I've said many times, some sites may still be experiencing it, independent of your knowledge or personal experience.
T
"If a tree in a forest falls and Peter is not there, does it make a sound?" -- Anonymous
On Tuesday, December 8, 2015 at 12:27:39 PM UTC-8, Peter McMurray wrote:
Similarly, when companies leave a platform they usually don't advertise the fact, burning bridges and embarrassing themselves with a public rant on their way out the door. They sometimes don't even tell their vendor, who may be made aware of a migration only when the Support renewal is not processed.
The fact that you did not know about this should not trouble you. Nor should you feel inclined to dismiss the severity, with the personal good fortune of not having experienced this yourself. Several people in this group have experienced the described problem - it's up to each individual to decide what they publish. And yes, many sites Have migrated, either from D3 Windows to Linux, or (dumped the product and) moved completely away from D3. This was not in the distant past, and was only fixed within the last few years ... and as I've said many times, some sites may still be experiencing it, independent of your knowledge or personal experience.
T
"If a tree in a forest falls and Peter is not there, does it make a sound?" -- Anonymous
On Tuesday, December 8, 2015 at 12:27:39 PM UTC-8, Peter McMurray wrote:
It seems that there is a major error history that I have never met. If people were losing 250Mb at any time that would be enough to dump the product. ...The major error must have been in the far distant past and long since fixed.
Charlie Noah
Dec 8, 2015, 10:41:38 PM12/8/15
to mvd...@googlegroups.com
Hey Tony,
I have to disagree with you on one point. In the ecommerce world, most companies have either built or modified their own websites, and usually find better and more prompt help on the forums than from their provider's support channel. It's sad, but true. I see it every day. Believe me, if a customer is unhappy enough to leave an ecommerce platform, they aren't usually quiet about it. Before you scold me, I do realize this is a different world than companies becoming dissatisfied with a Multivalue platform and leaving for what they think are greener pastures. And when they find out that grass isn't greener, they are even more silent. The end result is we all lose, and that's not a good thing.
Personally, I subscribe to this list, the jBASE Google Group, u2-users and even comp.databases.pick, although there isn't much there anymore. I get posts via email, and reply the same way. I generally don't go to their websites. But, that's just me. Different strokes, I guess! :-)
I think this thread has gone way off track, but if you're still reading, thanks.
I wish everyone a happy and blessed
Christmas
Bodhi Day
Winter Solstice
Chanukah
Id al-Fitr
Eid al-Adha
Ashura
Winter Solstice
Saturnalia
Yule
Festivus
Kwanzaa
Omisoka
Shabe-Yalda
Yuletide Season
Holidays
in no particular order (except my favorite), and I probably still missed some. If so, please don't take offense.
Charlie
I have to disagree with you on one point. In the ecommerce world, most companies have either built or modified their own websites, and usually find better and more prompt help on the forums than from their provider's support channel. It's sad, but true. I see it every day. Believe me, if a customer is unhappy enough to leave an ecommerce platform, they aren't usually quiet about it. Before you scold me, I do realize this is a different world than companies becoming dissatisfied with a Multivalue platform and leaving for what they think are greener pastures. And when they find out that grass isn't greener, they are even more silent. The end result is we all lose, and that's not a good thing.
Personally, I subscribe to this list, the jBASE Google Group, u2-users and even comp.databases.pick, although there isn't much there anymore. I get posts via email, and reply the same way. I generally don't go to their websites. But, that's just me. Different strokes, I guess! :-)
I think this thread has gone way off track, but if you're still reading, thanks.
I wish everyone a happy and blessed
Christmas
Bodhi Day
Winter Solstice
Chanukah
Id al-Fitr
Eid al-Adha
Ashura
Winter Solstice
Saturnalia
Yule
Festivus
Kwanzaa
Omisoka
Shabe-Yalda
Yuletide Season
Holidays
in no particular order (except my favorite), and I probably still missed some. If so, please don't take offense.
Charlie
Tiny Bear Wild Bird Store
Home of "Safety Net Shipping"
http://www.TinyBearWildBirdStore.com
Toll Free: 1-855-TinyBear (855-846-9232)
--
You received this message because you are subscribed to
the "Pick and MultiValue Databases" group.
To post, email to: mvd...@googlegroups.com
To unsubscribe, email to: mvdbms+un...@googlegroups.com
For more options, visit http://groups.google.com/group/mvdbms
Tony Gravagno
Dec 10, 2015, 9:14:30 PM12/10/15
to Pick and MultiValue Databases
Hi bud. We can certainly agree that there are no absolutes - some people post their concerns in forums and others do not.
Summary: People in forums usually can't resolve truly serious issues, so a forum isn't the right place for such queries. And serious decision makers aren't waiting for someone to post a free forum answer for serious issue - this isn't the right place for those people. Don't read further unless you have time to kill.
Here was my quote for reference:
Summary: People in forums usually can't resolve truly serious issues, so a forum isn't the right place for such queries. And serious decision makers aren't waiting for someone to post a free forum answer for serious issue - this isn't the right place for those people. Don't read further unless you have time to kill.
Here was my quote for reference:
"When companies have serious issues, they usually don't take them to forums. They contact their vendor and do what they can to collaborate toward a resolution."
If nothing else, my intent there was that the people who are most qualified to solve solutions are not in forums. You're not going to get an engineering team solving issues in an open discussion. We almost never see engineers or support people from any of the MVDBMS provider companies - and for good reason if my own history serves as any example. Even Martin Phillips will ask people to put in an official support call if a problem goes beyond basic product usage enquiries.
So if someone has a non-critical issue, it makes sense to see if someone else in the community has the same issue, but it doesn't make sense to hang around here waiting for some engineer to respond to a "serious issue" because the solution will not be found here. To be clear, losing 250MB of frames on a weekly basis is a serious issue, and everyone I know who has had that problem has taken it to their vendor.
Historically even in MV, we have a small, vocal group of passionate people. Similar types are found in most forums. But we don't see very much at all from the real decision makers, the C-level execs. People in forums "usually" are recommending technologies to decision makers, and/or implementing technical solutions. The few business owners we see in forums happen to have smaller shops and they happen to be more technical as well.
Not to present a challenge, but to clarify my intent: I doubt that anyone here could find a single business owner with more than 10 employees complaining in a forum about anything. These folks have first degated responsibility to employees to get a problem fixed. That process may start in a forum but then it must escallate directly through proper channels to the vendor. If not resolved, the next step of the business owner is to get on the phone to their vendor counterparts.
That applies to MV VARs as well, who are being paid by their end-users to contact the upline support provider on their behalf - not to rag in forums. When tech support doesn't work, it's time to escallate to management - not to rag in forums. Serious inter-business exchanges rarely happen in public, but we certainly see some amount of venting about what's happing in private. And with that, I suggest that we never have and never will see the owner of a company claiming in a forum that he/she is going to take business elsewhere due to some technical issue. It just doesn't happen. What we do see here are VARs who have taken those steps, from forum to support to management, and after continued dis-satisfaction they announce that they've migrated their clients to another platform from the same or a different vendor.
All of this comes from Peter's implication that excessive frame loss is rare, old, or possibly imagined, because there hasn't been great wailing and knashing of teeth in this forum. For many reasons I'm just saying (once again) he comes to many wrong conclusions for many wrong reasons.
Best,
T
So if someone has a non-critical issue, it makes sense to see if someone else in the community has the same issue, but it doesn't make sense to hang around here waiting for some engineer to respond to a "serious issue" because the solution will not be found here. To be clear, losing 250MB of frames on a weekly basis is a serious issue, and everyone I know who has had that problem has taken it to their vendor.
Historically even in MV, we have a small, vocal group of passionate people. Similar types are found in most forums. But we don't see very much at all from the real decision makers, the C-level execs. People in forums "usually" are recommending technologies to decision makers, and/or implementing technical solutions. The few business owners we see in forums happen to have smaller shops and they happen to be more technical as well.
Not to present a challenge, but to clarify my intent: I doubt that anyone here could find a single business owner with more than 10 employees complaining in a forum about anything. These folks have first degated responsibility to employees to get a problem fixed. That process may start in a forum but then it must escallate directly through proper channels to the vendor. If not resolved, the next step of the business owner is to get on the phone to their vendor counterparts.
That applies to MV VARs as well, who are being paid by their end-users to contact the upline support provider on their behalf - not to rag in forums. When tech support doesn't work, it's time to escallate to management - not to rag in forums. Serious inter-business exchanges rarely happen in public, but we certainly see some amount of venting about what's happing in private. And with that, I suggest that we never have and never will see the owner of a company claiming in a forum that he/she is going to take business elsewhere due to some technical issue. It just doesn't happen. What we do see here are VARs who have taken those steps, from forum to support to management, and after continued dis-satisfaction they announce that they've migrated their clients to another platform from the same or a different vendor.
All of this comes from Peter's implication that excessive frame loss is rare, old, or possibly imagined, because there hasn't been great wailing and knashing of teeth in this forum. For many reasons I'm just saying (once again) he comes to many wrong conclusions for many wrong reasons.
Best,
T
On Tuesday, December 8, 2015 at 7:41:38 PM UTC-8, CWNoah wrote:
Hey Tony,
I have to disagree with you on one point. In the ecommerce world, most companies ...usually find better and more prompt help on the forums than from their provider's support channel.
fwinans
Dec 22, 2015, 10:54:28 AM12/22/15
to Pick and MultiValue Databases
Erm, I think the (i option flag also nukes the frames in PEQS file.
So any printer jobs you had lurking in there, like perhaps some
big held reports you crank out overnight, will bump up that frame count.
So any printer jobs you had lurking in there, like perhaps some
big held reports you crank out overnight, will bump up that frame count.
fwinans
Dec 22, 2015, 4:08:02 PM12/22/15
to Pick and MultiValue Databases
You get the lost frames back if you remember to 'format' the vme area before doing the d3vme.exe /restore
-- replacing the disk0.d3v file in windows folder d3virtual is done by running disk0.exe from d3programs\ in
some versions of d3/nt, or unzipping disk0.zip from d3programs\ in other versions of d3. Check my spelling
on those, am shooting from the hip on a d3/linux box right now... This 'formatting' is not alluded to in the
rocket page detailing the d3vme.exe command line options,
http://www3.rocketsoftware.com/rocketd3/support/documentation/d3nt/91/admin/server_command_line_options.htm
which says in part that d3vme /restore
Clears the VME and rebuilds it with the data on the tape.
FSI accounts are overlayed if they exist on disk.
But I'm sure they mention it in various other docs. On d3/linux doing a full file restore implicitly formats
the frames for you, and a d3 service type person in CA caught me carelessly omitting this step once,
some years ago. Oops.
Frank
-- replacing the disk0.d3v file in windows folder d3virtual is done by running disk0.exe from d3programs\ in
some versions of d3/nt, or unzipping disk0.zip from d3programs\ in other versions of d3. Check my spelling
on those, am shooting from the hip on a d3/linux box right now... This 'formatting' is not alluded to in the
rocket page detailing the d3vme.exe command line options,
http://www3.rocketsoftware.com/rocketd3/support/documentation/d3nt/91/admin/server_command_line_options.htm
which says in part that d3vme /restore
Clears the VME and rebuilds it with the data on the tape.
FSI accounts are overlayed if they exist on disk.
But I'm sure they mention it in various other docs. On d3/linux doing a full file restore implicitly formats
the frames for you, and a d3 service type person in CA caught me carelessly omitting this step once,
some years ago. Oops.
Frank
Peter McMurray
Dec 22, 2015, 4:29:01 PM12/22/15
to Pick and MultiValue Databases
Hi Frank
I have never heard of using disk0 and it is not mentioned in the current manual - I am one of those weird people who reads everything before I turn the machine on.
However I see the program sitting there in the install. It would be good to see a response from Rocket with particular reference to Windows 10.
fwinans
Dec 22, 2015, 7:37:52 PM12/22/15
to Pick and MultiValue Databases
It took me a while to troll back through all my notes, but replacing disk0.d3v as part of doing a d3 / windows full vme restore
is mentioned way way back in TA 20 which deals with reducing the size of the vme {file NTTA20.PDF} See the original
Pick Systems logo at top of this document, dated Sep 11, 1997 {That was d3/nt on win95 way back then!}
Then came a May 13 2005 posting in the vendor-hosted discussion forum thread "VME Size Recommendations"
by Robert Burke {navigational subtitle of page is
Rocket Software Discussion Forums > Multi-Value Database Management Systems > D3 Windows
url of that posting is
http://d3mvforums.rocketsoftware.com/lofiversion/index.php?t827.html
but I found it easily by googling to search string d3 disk0.zip disk0.exe
I could be persuaded that manually needing to replace the disk0.d3v file before doing a d3vme /install
is a thing of the past, but not until somebody authoritatively tells so.
You wouldn't expect anything not done at the tcl : prompt to be in the otherwise delightful reference manual,
so things like howto do a full restore and what are the options for d3vme.exe aren't in the refman.
A quick search of the 9.2 installation guide doesn't find disk0 or vme
No joy googling with string site:www.d3ref.com and also the word vme or disk0
{That is the site that Zumasys puts up} hoping for a recent citation on this subject...
By the way, the watershed for zip versus executable self unpacking archive disk0 was
d3/nt 7.2.0 was disk0.zip then d3/nt 7.4.0 it became disk0.exe instead...
Although it is a bit scary running a .exe file somebody hands you, it is worth it not having to explain to an
end user out in the field they need to download an old pkzip dos shareware to unpack a provided zip; never
a fun phone call... I always wondered how the d3 install on same cdrom managed that unzip part...
I've was always fond of 7.2.0; it was the last release you could do a spurious d3 install without
performing an activation, which was really nice when doing cut/paste of the onscreen text seen during same
for in-house documentation, or just to get a feel that the install won't barf when you drive out to the client site
the next day because of something unexpected like unrecognizable os release or whatever.
Hint: when you
do an install of the base foo.zero d3 release and it demands an activation, then you instantly do an install of
the foo.1 patch release on top of that and _it_ demands an activation, it seems like you can just click on OK the
first time at the activation screen without bothering to activate; at least I think I noticed it that way once...
Frank
On Tuesday, December 22, 2015 at 3:29:01 PM UTC-6, Peter McMurray wrote:
is mentioned way way back in TA 20 which deals with reducing the size of the vme {file NTTA20.PDF} See the original
Pick Systems logo at top of this document, dated Sep 11, 1997 {That was d3/nt on win95 way back then!}
Then came a May 13 2005 posting in the vendor-hosted discussion forum thread "VME Size Recommendations"
by Robert Burke {navigational subtitle of page is
Rocket Software Discussion Forums > Multi-Value Database Management Systems > D3 Windows
url of that posting is
http://d3mvforums.rocketsoftware.com/lofiversion/index.php?t827.html
but I found it easily by googling to search string d3 disk0.zip disk0.exe
I could be persuaded that manually needing to replace the disk0.d3v file before doing a d3vme /install
is a thing of the past, but not until somebody authoritatively tells so.
You wouldn't expect anything not done at the tcl : prompt to be in the otherwise delightful reference manual,
so things like howto do a full restore and what are the options for d3vme.exe aren't in the refman.
A quick search of the 9.2 installation guide doesn't find disk0 or vme
No joy googling with string site:www.d3ref.com and also the word vme or disk0
{That is the site that Zumasys puts up} hoping for a recent citation on this subject...
By the way, the watershed for zip versus executable self unpacking archive disk0 was
d3/nt 7.2.0 was disk0.zip then d3/nt 7.4.0 it became disk0.exe instead...
Although it is a bit scary running a .exe file somebody hands you, it is worth it not having to explain to an
end user out in the field they need to download an old pkzip dos shareware to unpack a provided zip; never
a fun phone call... I always wondered how the d3 install on same cdrom managed that unzip part...
I've was always fond of 7.2.0; it was the last release you could do a spurious d3 install without
performing an activation, which was really nice when doing cut/paste of the onscreen text seen during same
for in-house documentation, or just to get a feel that the install won't barf when you drive out to the client site
the next day because of something unexpected like unrecognizable os release or whatever.
Hint: when you
do an install of the base foo.zero d3 release and it demands an activation, then you instantly do an install of
the foo.1 patch release on top of that and _it_ demands an activation, it seems like you can just click on OK the
first time at the activation screen without bothering to activate; at least I think I noticed it that way once...
Frank
On Tuesday, December 22, 2015 at 3:29:01 PM UTC-6, Peter McMurray wrote:
I have never heard of using disk0 and it is not mentioned in the current manual.
Tony Gravagno
Dec 23, 2015, 5:45:43 AM12/23/15
to mvd...@googlegroups.com
I've been working with this stuff every day for 20 years. I worked in Support. I managed QA. I was the product manager. And I have no idea what you're talking about.
Your D3 Linux is completely different and unrelated to the topic.
d3ref.com was an unauthorized reproduction of old documentation by Sierra-Bravo. Raining Data chose not to request a take down, despite my suggestion then, and more valid now, that it does them and their Current client base a huge disservice. I believe the agreed remedy was to put a copyright footer on the bottom of all pages. The site was passed on to Nerdery and then to Zumasys without updates except for their own branding, as if it's maintained. It remains a dead and misleading archive for anyone with a current system. The insane copyright notice is updated yearly, only lending more misleading credibility to the 19 year old content which is apparently still referenced and quoted by "value-add" resellers.
The current docs are on the Rocket site.
T
Peter McMurray
Dec 23, 2015, 8:29:21 PM12/23/15
to Pick and MultiValue Databases
Hi
The current Rocket disk0.exe has the following help
Rocket D3 Disk 0
Command Line Options
?Q -- Quiet Mode for Packages
/T:<full path> -- Specifies temporary working folder
/C -- Extract files only to the folder when used also with ?T.
?c:<Cmd> -- Override Install Command defined by Author
:which
System Release Information
==========================
D3 Release Version 9.2.1.WINDOWS
Implementation. . . . . . WINDOWS
Software Serial Number. . 11031377
System ID Number . . . . 70348075
Release . . . . . . . . . D3
Windows Information . . . Windows;pick0;6.2.9200;;6.2
:Windows 10
The thought of trying to get a user to unzip something and install anything never mind something as important as D3 is beyond my wildest nightmares. I recently had someone type in c:\newfile as c:`newfile proving that they cannot even copy from an email. The grave accent does slope back I suppose :-( I discovered this when their wretched Cisco VPN system finally allowed me in..
fwinans
Dec 24, 2015, 2:39:40 PM12/24/15
to Pick and MultiValue Databases
disk0.exe is a self-unpacking archive; running it puts up a dialog box asking your
desires as to location of the resulting disk0.d3v file.
The compression is really worthwhile; on d3/nt 7.4.0 the disk0.exe was 2.647 megs
versus the resulting disk0.d3v was initially 32.25 megabytes {when you next fire up
d3vme.exe it will enlarge that 32 megs to the size of the vme as seen in your d3
device manager applet.}
The question I've had is just how energetic d3vme.exe /restore is about cleaning
house in this file; if I'm wanting to make sure no accrued gfe's survive, should I bother
replacing the file from the disk0.exe beforehand? I see Tony posted a reply, but it
was a bit too broad to make plans from. I'll just carry on as before, as there's no harm
in it, and it may help sometimes. I'm not going to bother filing a support case,
there's not real gain in making even that slight effort on such a minor matter.
desires as to location of the resulting disk0.d3v file.
The compression is really worthwhile; on d3/nt 7.4.0 the disk0.exe was 2.647 megs
versus the resulting disk0.d3v was initially 32.25 megabytes {when you next fire up
d3vme.exe it will enlarge that 32 megs to the size of the vme as seen in your d3
device manager applet.}
The question I've had is just how energetic d3vme.exe /restore is about cleaning
house in this file; if I'm wanting to make sure no accrued gfe's survive, should I bother
replacing the file from the disk0.exe beforehand? I see Tony posted a reply, but it
was a bit too broad to make plans from. I'll just carry on as before, as there's no harm
in it, and it may help sometimes. I'm not going to bother filing a support case,
there's not real gain in making even that slight effort on such a minor matter.
On Wednesday, December 23, 2015 at 7:29:21 PM UTC-6, Peter McMurray wrote:
The current Rocket disk0.exe has the following helpRocket D3 Disk 0Command Line Options?Q -- Quiet Mode for Packages/T:<full path> -- Specifies temporary working folder/C -- Extract files only to the folder when used also with ?T.?c:<Cmd> -- Override Install Command defined by Author
<snip>
fwinans
Dec 24, 2015, 4:00:24 PM12/24/15
to Pick and MultiValue Databases
Oops. Both of those places describe doing a replace of file disk0.d3v
but for a totally different reason than why I tend to do it;
they're all about reducing the vme size, and I'm all about making sure any
past gfe's are dead and buried, even without any firm assertions about
the various neat things d3vme /restore may or may not do beyond sucking
tape onto vme. Sorry about that fuzzy thinking on my part.
On Tuesday, December 22, 2015 at 6:37:52 PM UTC-6, fwinans wrote:
but for a totally different reason than why I tend to do it;
they're all about reducing the vme size, and I'm all about making sure any
past gfe's are dead and buried, even without any firm assertions about
the various neat things d3vme /restore may or may not do beyond sucking
tape onto vme. Sorry about that fuzzy thinking on my part.
On Tuesday, December 22, 2015 at 6:37:52 PM UTC-6, fwinans wrote:
<snip> replacing disk0.d3v as part of doing a d3 / windows full vme restore
is mentioned way way back in TA 20 which <snip>
<snip>
Then came a May 13 2005 posting <snip>
http://d3mvforums.rocketsoftware.com/lofiversion/index.php?t827.html
<snip>
Mark Brown
Feb 2, 2016, 4:58:30 AM2/2/16
to Pick and MultiValue Databases
In 1991 Dave Zigray asked me to re-write the spooler to make better use of open architecture and add features and functionality. There was some pretty neat stuff, but for some reason they dropped most of it when they dropped the R/T product line.
But, there is a param frame with pointers to everything and one of those is the Spooler Control Block (SCB). That contain a table of printer entries and a table of spooler entries. The printer entries point to frames containing the printer control block (PCB), the address of the current jobs being output, and control and status bits; Each print element has an entry in the table and at least 1 frame of overflow used for the output.
Sometimes those pointers get out of sync. The technical term is screwed up.
STARTSPOOLER (C will clear some of the control bits in the two tables but pretty much leave everything alone. STARTSPOOLER (I zeros that original params frame pointer and you loose the entire spooler tree. There could easily be a quarter GB of data tossed away with a single command
That's why Jon said, last resort.
Reply all
Reply to author
Forward
0 new messages