We have turned off our own news servers (well, turned off one and the other
turned itself off), and pointed news.hiwaay.net to the [bcandid.com] news
service, and the newstoo.hiwaay.net to the [supernews.com] service.
...each of the services claims theirs
is better.) Try them both, and let us know which YOU feel is better.
email sup...@hiwaay.net with your opinion.
________________________
I have not used Hiwaay News in two since I've been away at school. I'm
using BellSouth News Servers and they seem to work pretty good. Only one
outtage in two years. Dunno if I'll want to try Hiwaay News when I get
back to Huntsville in January.
Jesse
>We have turned off our own news servers (well, turned off one and the other
>turned itself off), and pointed news.hiwaay.net to the [bcandid.com] news
>service, and the newstoo.hiwaay.net to the [supernews.com] service.
They had pointed it into the trash! I tried for 20 minutes to get into the
news.
>...each of the services claims theirs
>is better.) Try them both, and let us know which YOU feel is better.
>email sup...@hiwaay.net with your opinion.
>________________________
>
After 30 minutes on hold trying to speak to someone who believes "your call
is important to us" I gave up.
Maybe all of us Hiwaay users should "point our wallets" at another ISP.
What are some good ISP options in N. Alabama?
Respectfully,
Mike
I like Mindspring, they give you 30 days free to start. Try them and if you do
not like them, drop them.
Rex
Jesse
In article <7oeits$3th$1...@hammer.msfc.nasa.gov>, "Mike Travis"
Jesse wrote:
> HiWaay Says:
>
> We have turned off our own news servers (well, turned off one and the other
> turned itself off), and pointed news.hiwaay.net to the [bcandid.com] news
> service, and the newstoo.hiwaay.net to the [supernews.com] service.
Ah, that's why HiWaay's news service has been abysmally slow lately. These new
servers suck compared to the old setup.
jsp
Well, it was sure nice if them to let us know beforehand!!!
Bill L.
Irate!
#######################################################################
bi...@Traveller.COM http://www.hsv.tis.net/~bill
He that is of a merry heart hath
a continual feast - Prov. 15:15b
>jess...@nospam.yahoo.com (Jesse) wrote:
>>HiWaay Says:
>>
>>We have turned off our own news servers (well, turned off one and the other
>>turned itself off), and pointed news.hiwaay.net to the [bcandid.com] news
>>service, and the newstoo.hiwaay.net to the [supernews.com] service.
>>
>>....each of the services claims theirs
>>is better.) Try them both, and let us know which YOU feel is better.
>>email sup...@hiwaay.net with your opinion.
>Well, it was sure nice if them to let us know beforehand!!!
Really. We (news.msfc.nasa.gov) exchange a small feed with one of their
machines, and the only way I knew something was wrong was that I had a
20,000 article backlog going to their site.
It's okay, though, no hard feelings...
--
J. Porter Clark, d/b/a
+--+
|oo| The Unknown News Administrator
| | ne...@news.msfc.nasa.gov
`^^'
supernews.com (newstoo): slow, often quits for many tens of seconds
while reading images from alt.binaries.3d.bryce, often several
times per image. Sometimes never finishes the download. Hangs
when trying to send posts (3 times so far), will not send for
several minutes. Only cure is cancel and push send again. Guess
I'll find out if there are duplicate postings and the news server
just failed to complete the transaction, or whether it never got the
post in the first place.
bcandid.com (news): unbelievably, outrageously slow. Even worse on
binaries than supernews. Have not tried to send on it yet.
So far, not good. In fact, news.povray.org, which used to be my
standard for slow -- so much so that I'd go several days without
connecting because of the aggravation -- is faster than either of
the new news hosts.
--
Bob Crispen
crispen at hiwaay dot net
What we're looking for: destinations.
What we end up getting: journeys.
So, what happened to the glorious InterGraph NT "bulletproof" ultra-
reliable server? Did the hardware lose its smoke, or did NT commit
seppu-ku when faced with a real job? Hugh? Hugh? Anybody? Bueler?
>> ...each of the services claims theirs
>> is better.) Try them both, and let us know which YOU feel is better.
>> email sup...@hiwaay.net with your opinion.
My relatively informal testing indicates they both suck. See traceroutes
below...
>supernews.com (newstoo): slow, often quits for many tens of seconds
> while reading images from alt.binaries.3d.bryce, often several
> times per image. Sometimes never finishes the download. [ ... ]
Never finishes a traceroute, either. Granted, that could be an artifact
of their network configuration, but the times don't make any of it look
good.
>bcandid.com (news): unbelievably, outrageously slow. Even worse on
> binaries than supernews. Have not tried to send on it yet.
[ ... ]
A traceroute to each is below. They were run at about 1830 this evening.
Perhaps HiWAAY should consider a caching server...
BTW, this brings to mind another question--has the Huntsville area ISP
interconnection been dissolved? The path between ro.com and hiwaay.net
now loops through Atlanta, whereas previously it did not. Did the recent
consolidations cause the plug to be pulled?
Gary
This trace was to news.hiwaay.net:
traceroute to ga231beo.wip.ispnews.com (206.132.58.118), 30 hops max, 40 byte packets
1 Rogate (205.216.92.1) 2.063 ms 2.036 ms 2.372 ms
2 border3-serial2-3.Atlanta.cw.net (204.70.18.181) 16.989 ms 22.623 ms 25.167 ms
3 core1-fddi-0.Atlanta.cw.net (204.70.2.49) 15.456 ms 19.236 ms 42.454 ms
4 core2.Sacramento.cw.net (204.70.4.49) 99.349 ms 90.754 ms 98.451 ms
5 border8-fddi-0.Sacramento.cw.net (204.70.164.67) 83.999 ms 82.997 ms 85.190 ms
6 globalcenter.Sacramento.cw.net (204.70.123.6) 81.543 ms 75.954 ms 84.997 ms
7 pos4-3-155M.cr1.SNV.globalcenter.net (206.132.150.29) 82.087 ms 114.426 ms 81.308 ms
8 pos2-0-622M.cr1.SNV2.globalcenter.net (206.132.151.9) 87.613 ms 87.832 ms 75.743 ms
9 pos11-0-0-155M.hr1.SNV2.globalcenter.net (206.132.151.42) 76.117 ms 277.196 ms 88.461 ms
10 host118.misi.com (206.132.58.118) 111.699 ms 76.889 ms 82.144 ms
This was to newstoo.hiwaay.net:
traceroute to corp.supernews.com (207.126.101.100), 30 hops max, 40 byte packets
1 Rogate (205.216.92.1) 2.329 ms 1.938 ms 2.216 ms
2 border3-serial2-3.Atlanta.cw.net (204.70.18.181) 10.808 ms 11.035 ms 11.864 ms
3 core1-fddi-0.Atlanta.cw.net (204.70.2.49) 10.378 ms 21.870 ms 13.736 ms
4 core4.WillowSprings.cw.net (204.70.4.73) 36.554 ms 31.359 ms 39.799 ms
5 aads.above.net (198.32.130.71) 33.492 ms 33.973 ms 39.073 ms
6 chicago-core1.ord.above.net (209.249.0.129) 60.703 ms 31.162 ms 34.768 ms
7 sjc-ord-oc12.sjc2.above.net (207.126.96.118) 208.648 ms 210.073 ms 204.055 ms
8 core5-core1-oc48.sjc.above.net (216.200.0.177) 163.235 ms 192.324 ms 197.382 ms
9 main1-core5.sjc.above.net (209.133.31.134) 180.600 ms 208.499 ms 198.582 ms
10 * * *
[ sigh... ]
29 * * *
30 * * *
--
Gary Heston ghe...@ro.com spamfodder: root@localhost pres...@whitehouse.gov
sync@localhost daemon@localhost lp@localhost mail@localhost news@localhost
"CHKDSK.EXE - Corrupt File The file or directory \$Mft is corrupt or
unreadable. Please run the Chkdsk utility." NT Server 4.0 error message
I have it from a reliable source that they were trying out some fibre
channel stuff. This may have bit it. The News Admin is an old UNIX(r)
guy, so it's for ease of use that HiWAAY runs NT.
> BTW, this brings to mind another question--has the Huntsville area ISP
> interconnection been dissolved? The path between ro.com and hiwaay.net
> now loops through Atlanta, whereas previously it did not. Did the recent
> consolidations cause the plug to be pulled?
My guess is yes.
--
Kris Kirby
<kr...@airnet.net>
-------------------------------------------
TGIFreeBSD... 'Nuff said.
No, nothing has changed in the local ISP peering (except the number of
ISPs). From sh1.ro.com:
$ traceroute www.hiwaay.net
traceroute: Warning: Multiple interfaces found; using 205.216.92.5 @ eth0
traceroute to A.RaQ1.HiWAAY.net (208.170.109.3), 30 hops max, 40 byte packets
1 Rogate (205.216.92.1) 3.421 ms 2.281 ms 5.013 ms
2 smith-broadcasting.Atlanta.cw.net (166.48.45.250) 26.006 ms 10.220 ms 9.128 ms
3 a.raq1.hiwaay.net (208.170.109.3) 11.574 ms 10.054 ms 10.327 ms
$
The "smith-broadcasting.Atlanta.cw.net" is actually HiWAAY's router.
The "cw.net" hostname is an artifact of the way HiWAAY does routing with
unnumbered interfaces, so their router just picks an IP address from a
numbered interface (usually an ethernet) to use (and it has picked their
interface to CWIX, since CWIX always uses numbered interfaces).
If we didn't have the peering, the first hop past our router (Rogate)
would be either border3-serial2-3.Atlanta.cw.net (where RO/IBS connects
to CWIX) or dca01.atm-e.us.crl.net (where RO/IBS connects to CRL). It
would typically be the first one since HiWAAY is also connected to CWIX
and that would be the preferred route.
--
Chris Adams <cad...@ro.com> - System Administrator
Renaissance Internet Services - IBS Interactive, Inc.
Home: http://ro.com/~cadams - Public key: http://ro.com/~cadams/pubkey.txt
I don't speak for anybody but myself - that's enough trouble.
>Maybe all of us Hiwaay users should "point our wallets" at another ISP.
>What are some good ISP options in N. Alabama?
>
>Respectfully,
>Mike
Hiwaay bought the other local ISP's [We are Hiwaay. You will be
assimilated. Resistance is futile]. There is Mindspring, which a friend of
mine in South Carolina uses. I've used hers and it was pretty good. I
don't really know about Bellsouth.net.
--
Bradley Lemmond
Unencumbered by the thought process
Remove X and Y when replying
http://www.airnet.net/blemmond/
If you want to know what God thinks about money, just look at the sort of
people he's given it to.
--Audrey Forbes-Hamilton, "To the Mannor Born" BBC
Mike Travis wrote in message <7oeits$3th$1...@hammer.msfc.nasa.gov>...
> Well, it was sure nice if them to let us know beforehand!!!
I talked to the tech support staff, and they promise that in
the future their machines will only break when they're
scheduled to.
I can only assume that means they're switching over to NT.
;-)
Actually, I can't get my knickers in too much of a twist over
this. The folks at Hiwaay seem pretty driven by quality,
and the level of performance we're getting out of the new
news servers must irk them even more than it does us.
But for goodness sakes, in case there's an MBA hanging around
in the Hiwaay offices, do send in your complaints if you have
any (compliments as well). They won't officially know unless
we tell them.
Btw, the server on newstoo did pick up the hiwaay groups today,
so maybe we should take this discussion there.
>>My guess is yes.
>No, nothing has changed in the local ISP peering (except the number of
>ISPs).
Then something strange is going on.
> From sh1.ro.com:
>$ traceroute www.hiwaay.net
>traceroute: Warning: Multiple interfaces found; using 205.216.92.5 @ eth0
>traceroute to A.RaQ1.HiWAAY.net (208.170.109.3), 30 hops max, 40 byte packets
> 1 Rogate (205.216.92.1) 3.421 ms 2.281 ms 5.013 ms
> 2 smith-broadcasting.Atlanta.cw.net (166.48.45.250) 26.006 ms 10.220
>ms 9.128 ms
> 3 a.raq1.hiwaay.net (208.170.109.3) 11.574 ms 10.054 ms 10.327 ms
And when I do a traceroute from work to sh1.ro.com, I get this:
Tracing route to sh1.ro.com [205.216.92.5]
over a maximum of 30 hops:
[ work stuff deleted ]
4 * 167 ms 134 ms hsvrouter.hiwaay.net [208.147.154.33]
5 220 ms 134 ms 139 ms bordercore1-hssi0-0-0.atlanta.cw.net [166.48.45.249]
6 158 ms 140 ms 142 ms core1.atlanta.cw.net [204.70.4.141]
7 206 ms 138 ms 146 ms border3-fddi-0.atlanta.cw.net [204.70.2.51]
8 675 ms 252 ms * phase-iv.atlanta.cw.net [204.70.18.182]
9 159 ms 156 ms 151 ms sh1.ro.com [205.216.92.5]
As far as I can tell, this is *not* a local connection path.
>The "smith-broadcasting.Atlanta.cw.net" is actually HiWAAY's router.
>The "cw.net" hostname is an artifact of the way HiWAAY does routing with
>unnumbered interfaces, so their router just picks an IP address from a
>numbered interface (usually an ethernet) to use (and it has picked their
>interface to CWIX, since CWIX always uses numbered interfaces).
Interesting, but it adds to confusion.
>If we didn't have the peering, the first hop past our router (Rogate)
>would be either border3-serial2-3.Atlanta.cw.net (where RO/IBS connects
>to CWIX) or dca01.atm-e.us.crl.net (where RO/IBS connects to CRL). It
>would typically be the first one since HiWAAY is also connected to CWIX
>and that would be the preferred route.
See above. As I recall, the bordercore1-hssi0-0-0 connection is the
usual route-to-far-places hop. Something is not quite right here--
perhaps they're using the news server for maintaining the routing
tables, or something. This used to work right, through hsv.mae.net
or something like that.
Gary
What's their timeout set on their newservers? I don't
use a offline reader. Hiwaay's new newsservers drop off
before I can read some of the longer posts.
Dan Harper
Yup, your right. It looks like HiWAAY is missing our routes in their
routing tables. We send traffic to them, but they don't send any back
(asymmetric routing is fun!). I'll have to look in to this some more.
Actually, news.hiwaay.net and newstoo.hiwaay.net were both running NT and
had been since early '97 if I recall correctly. Basically, newsfeeds have
outgrown those servers or were getting to the point were major hardware
upgrades would be necessary so they did the math and decided to outsource.
Once I got my newsreader reconfigured the new servers seem to pretty good
and response is surprising good on news.hiwaay.net, I can't say I notice the
difference.
--
Nik Simpson
Well, if it was a broken newsserver, that's a bit different.
I apologize if I over-reacted.
Bill L.
news.hiwaay.net is kind of iffy. I just did a traceroute with
NeoTrace and it showed a huge string of hosts (many not responding)
between here and news. But other times over the weekend it was
faster than newstoo (not the brand-new newstoo).
We have been trying various services and seeing what the response from our
customers has been... the service at newstoo.hiwaay.net is starting to
look like a keeper, and the number of groups they carry is actually
greater than what we took on the old pet 'no binaries' machine... it takes
about 35,000 groups and looks to be well administered, after some bumpy
starts with them it's looking pretty good.... equal to or better than pet
was by all measures.
The service at news.hiwaay.net is the identical service used by two of our
major competitors that I would occassionaly hear about from a subscriber
making comparisons that it was so much better than HiWAAY's server. That
must be in the eye of the beholder.... it takes a lot of groups, about
50,000 but it's performance is not what we would like (although the
promise it will be better soon)....
Today we are going to try replacing news.hiwaay.net with an start-up
companies service that promises 65,000 groups (yep I figure about
50,000 of those groups will never ever be read even once by our
subscribers but if that's the metric then we'll do our best)....
=======================================================================
Mark Derrick, General Mgr HiWAAY Information Services
Telephone (256) 533-4296 721 Clinton Avenue, Suite #8
http://www.HiWAAY.net Huntsville, AL 35801
Be careful about that grass on the other side of the fence; being
greener doesn't necessariy mean better.
>We have been trying various services and seeing what the response from our
>customers has been... the service at newstoo.hiwaay.net is starting to
>look like a keeper, and the number of groups they carry is actually
>greater than what we took on the old pet 'no binaries' machine... it takes
>about 35,000 groups and looks to be well administered, after some bumpy
>starts with them it's looking pretty good.... equal to or better than pet
>was by all measures.
Have you considered using a caching server locally fed from this service?
They can be configured to only pull over a group when someone requests it,
then keeps it locally until it expires, so that if a second user (or a few
dozen) also want to see it, the articles are only pulled from the feed
site once instead of two (or a few dozen) times. I know of a (non-local)
site which does this, and it seems to work well. Everything is available,
but if it's not going to be read, it's not transferred.
>The service at news.hiwaay.net is the identical service used by two of our
>major competitors that I would occassionaly hear about from a subscriber
>making comparisons that it was so much better than HiWAAY's server. That
>must be in the eye of the beholder.... it takes a lot of groups, about
>50,000 but it's performance is not what we would like (although the
>promise it will be better soon)....
>Today we are going to try replacing news.hiwaay.net with an start-up
>companies service that promises 65,000 groups (yep I figure about
>50,000 of those groups will never ever be read even once by our
>subscribers but if that's the metric then we'll do our best)....
Sounds like a perfect situation for a caching server.
Gary
--
Gary Heston ghe...@ro.com spamfodder: root@localhost pres...@whitehouse.gov
sync@localhost daemon@localhost lp@localhost mail@localhost news@localhost
"The processes? Kill them all, and let INIT sort them out!"