Network routing problems have been fixed


log in

Advanced search

Message boards : News : Network routing problems have been fixed

Previous · 1 · 2 · 3 · Next
Author Message
Profile HAL9000
Volunteer tester
Avatar
Send message
Joined: 11 Sep 99
Posts: 3851
Credit: 106,825,833
RAC: 95,610
United States
Message 1165441 - Posted: 26 Oct 2011, 7:56:48 UTC

Did they ever figure out why there was suddenly not enough ram in the router? Possibly the release of IPv6 on the internet as the issues started about the same time IIRC.
____________
SETI@home classic workunits: 93,865 CPU time: 863,447 hours

Join the BP6/VP6 User Group today!

Profile Pato
Send message
Joined: 27 Nov 06
Posts: 4
Credit: 679,945
RAC: 151
Australia
Message 1165663 - Posted: 27 Oct 2011, 1:51:45 UTC

Hurray, I'm connecting again without going via a proxy! W00t!
____________

DrFoo
Send message
Joined: 17 Jul 99
Posts: 26
Credit: 23,005,832
RAC: 27,045
United States
Message 1166120 - Posted: 29 Oct 2011, 1:50:42 UTC

Wow, I don't think there's any question this has actually been effecting us all to a major degree. Work units are flying in here at 4 seconds each (pretty much where I have the cap set on my end), no delays, no timeouts, no having to retry over and over.

According to the graphs it looks like your bandwidth use has been almost cut in HALF. Nothing kills a network like a flaky router...

Kudos!
____________

msattler
Volunteer tester
Avatar
Send message
Joined: 9 Jul 00
Posts: 38315
Credit: 558,712,195
RAC: 628,274
United States
Message 1166121 - Posted: 29 Oct 2011, 1:57:41 UTC - in response to Message 1166120.

Wow, I don't think there's any question this has actually been effecting us all to a major degree. Work units are flying in here at 4 seconds each (pretty much where I have the cap set on my end), no delays, no timeouts, no having to retry over and over.

According to the graphs it looks like your bandwidth use has been almost cut in HALF. Nothing kills a network like a flaky router...

Kudos!

The bandwidth use has nothing to do with the router fix........
Right now there is a cap limit on cache size...so this is what things would be like if the project was up and stable long enough for all caches to be filled.
But mostly, right now, no AP work is being split and sent out.
AP takes a bit more, percentage wise, of the bandwidth per unit of crunching time by the hosts. So when AP is going out, it tends to saturate the bandwidth.

What the router fix DID do however, was allow almost everybody to consistently connect with the servers. Which had been a problem for quite some time.

Packet loss when the bandwidth is saturated has been attributed to 'load limiting' by the routers/converters/whatever boxen that are in the Seti link and limit bandwidth to the 100Mbit they are allowed to use.
____________
*********************************************
Embrace your inner kitty...ya know ya wanna!

I have met a few friends in my life.
Most were cats.

DrFoo
Send message
Joined: 17 Jul 99
Posts: 26
Credit: 23,005,832
RAC: 27,045
United States
Message 1166129 - Posted: 29 Oct 2011, 2:16:24 UTC - in response to Message 1166121.

The bandwidth use has nothing to do with the router fix........

Well, I'm not going to argue about it, but I'm not so sure. Seems to me the cap has been in place for a while now, but maybe not as long as I was thinking.

Both changes are certainly making an improvement, and things are running better right now than they have in a very long while. I guess that's all that matters.


____________

msattler
Volunteer tester
Avatar
Send message
Joined: 9 Jul 00
Posts: 38315
Credit: 558,712,195
RAC: 628,274
United States
Message 1166130 - Posted: 29 Oct 2011, 2:19:11 UTC - in response to Message 1166129.

The bandwidth use has nothing to do with the router fix........

Well, I'm not going to argue about it, but I'm not so sure. Seems to me the cap has been in place for a while now, but maybe not as long as I was thinking.

Both changes are certainly making an improvement, and things are running better right now than they have in a very long while. I guess that's all that matters.


Not meaning to argue...LOL.
Just trying to explain what is going on right now and why you are seeing such moderate bandwidth usage and good transfer rates right now.

Like you said, things are running pretty smoothly.

Meow!
____________
*********************************************
Embrace your inner kitty...ya know ya wanna!

I have met a few friends in my life.
Most were cats.

Jon Stoltz
Send message
Joined: 22 Jan 08
Posts: 6
Credit: 338,123
RAC: 0
United States
Message 1166901 - Posted: 1 Nov 2011, 4:57:16 UTC

question? as of, well i noticed on friday night. i have noticed my transfers are not going through well. i have to babysit them. they keep going "retry in xxxx" which is sometime hours. and othertimes, it says project back off..... what gives?

juan BFB
Volunteer tester
Avatar
Send message
Joined: 16 Mar 07
Posts: 4935
Credit: 268,913,079
RAC: 353,442
Brazil
Message 1166903 - Posted: 1 Nov 2011, 5:09:52 UTC

The same here...

Profile Gatekeeper
Avatar
Send message
Joined: 14 Jul 04
Posts: 887
Credit: 176,479,616
RAC: 30
United States
Message 1166912 - Posted: 1 Nov 2011, 7:01:47 UTC - in response to Message 1166901.

question? as of, well i noticed on friday night. i have noticed my transfers are not going through well. i have to babysit them. they keep going "retry in xxxx" which is sometime hours. and othertimes, it says project back off..... what gives?


200,000 computers all vying for a piece of the pie, bandwidth-wise. A VHAR storm based on the recent mix of available work. Re-introduction of AP unit distribution at 8MB a pop. Looks like, despite all that, work is starting to get out at a better rate, as of 0700 UTC
____________

DrFoo
Send message
Joined: 17 Jul 99
Posts: 26
Credit: 23,005,832
RAC: 27,045
United States
Message 1167065 - Posted: 2 Nov 2011, 0:22:22 UTC

Sorry everyone, I obviously jinxed it by mentioning how well everything was working! Oh well, just back to normal I guess. It's all moving, just not as fast as we'd like.
____________

Profile Mark St. Denis
Send message
Joined: 20 May 99
Posts: 23
Credit: 4,439,263
RAC: 4,670
United States
Message 1167067 - Posted: 2 Nov 2011, 0:32:43 UTC - in response to Message 1167065.

Not normal. Taking 2 hours to download 50 tasks.
____________

Profile Italy98
Avatar
Send message
Joined: 15 May 99
Posts: 27
Credit: 394,994
RAC: 0
United States
Message 1167447 - Posted: 3 Nov 2011, 11:29:24 UTC - in response to Message 1166901.

I started a new download last night and this morning I find all the processes are still in the transfer mode with the status backoff 10+ hours.

question? as of, well i noticed on friday night. i have noticed my transfers are not going through well. i have to babysit them. they keep going "retry in xxxx" which is sometime hours. and othertimes, it says project back off..... what gives?


____________

Who Know's
Send message
Joined: 12 Sep 99
Posts: 9
Credit: 9,285,433
RAC: 2,674
United States
Message 1168218 - Posted: 5 Nov 2011, 14:14:54 UTC - in response to Message 1167447.

hi,
well before the 4th nov. my problem was downloading them, now today the 6th nov. i am having problems with the upload of them

rob smith
Volunteer tester
Send message
Joined: 7 Mar 03
Posts: 8127
Credit: 52,577,681
RAC: 75,080
United Kingdom
Message 1168223 - Posted: 5 Nov 2011, 14:39:30 UTC

It looks more like a server side problem than the router.
Have a look at the "server status" page, its last update as 21:20UTS...
Also some discussion on the number crunchers thread.
____________
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

Jon Stoltz
Send message
Joined: 22 Jan 08
Posts: 6
Credit: 338,123
RAC: 0
United States
Message 1168347 - Posted: 5 Nov 2011, 20:28:47 UTC

i seem to download ok at times, i havent been able to upload. every great once in a while i can. but this is just a big mess.
____________

Nautilus
Avatar
Send message
Joined: 19 Jun 06
Posts: 44
Credit: 889,174
RAC: 0
Turkey
Message 1168607 - Posted: 6 Nov 2011, 11:28:58 UTC
Last modified: 6 Nov 2011, 11:29:30 UTC

I can't download any WUs...
____________

Richard Haselgrove
Volunteer tester
Send message
Joined: 4 Jul 99
Posts: 8375
Credit: 46,659,399
RAC: 18,242
United Kingdom
Message 1168608 - Posted: 6 Nov 2011, 11:32:31 UTC - in response to Message 1168607.

I can't download any WUs...

Sorry, but "user can't download" is not news.

Please discuss in Number Crunching, or read the diagnostics which other users have already posted there.

Roger Vaught
Send message
Joined: 3 Jun 01
Posts: 3
Credit: 384,769
RAC: 0
United States
Message 1176849 - Posted: 9 Dec 2011, 19:21:37 UTC

Just got back on BOINC after my 2 Macs were stolen. I installed BOINC on both new replacements using the default settings on both. On the Macbook, both Seti@home and Rosetta are working fine.

On iMac, both Seti@home and Rosetta show as Projects but only Rosetta is showing any tasks.

I haven't done anything different in the installations. Highlighting seti, hitting update gives the status Scheduler request pending Communication deferred for 5 minutes. Any suggestions how I can get Seti running on the iMac?

Thanks,
Roger Vaught

iMac 2.7 GHz Intel Core i5
Mmeory 4Gb
Graphics AMD Radeon HD 6770M 512 Mb
Lion 10.7.2

OzzFan
Volunteer tester
Avatar
Send message
Joined: 9 Apr 02
Posts: 13541
Credit: 29,356,099
RAC: 15,847
United States
Message 1176861 - Posted: 9 Dec 2011, 19:38:44 UTC - in response to Message 1176849.

Sorry to hear about your Macs being stolen. You can find plenty of help outside of this News forum at Questions and Answers, or some of the best minds can be found in Number Crunching.

Profile donaldjj
Send message
Joined: 30 Nov 00
Posts: 19
Credit: 1,345,310
RAC: 1,747
United States
Message 1176871 - Posted: 9 Dec 2011, 20:02:03 UTC - in response to Message 1176861.

Now to see if the thief is smart enough to turn off seti or just let it run in background and get tracked to their new homes :-)
____________

Previous · 1 · 2 · 3 · Next

Message boards : News : Network routing problems have been fixed

Copyright © 2014 University of California