Welcome to MilkyWay@home

Server Problems? - currently the U/L one

Message boards : Number crunching : Server Problems? - currently the U/L one
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 5803 - Posted: 2 Nov 2008, 7:58:52 UTC - in response to Message 5793.  

I don't think many will appear untill the next group is added. The total has gone from 36000 to 17700 wu's.


You are right there banditwolf.

The server status page, at 07.48 (and 11 seconds) UTC reports the stock of results in progress now down to 14,763.

Although I have been looking for, and waiting, the next batch of WUs, to give MW crunching priority here. I think the Admins intend to run down the results in progress first. That way they can release any new tweaked faster client and a new WU looking at more science and accuracy (longer WU)?
ID: 5803 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile the silver surfer
Avatar

Send message
Joined: 27 Nov 07
Posts: 39
Credit: 1,207,109
RAC: 0
Message 5806 - Posted: 2 Nov 2008, 10:21:56 UTC

Hi all !

Back to Milkyway, using Milksops` opt. app ( Thank You ! )
All went well ( except no new work, of course ), until I got
this message :
02.11.2008 11:01:16|Milkyway@home|Fetching scheduler list
02.11.2008 11:01:21|Milkyway@home|Deferring communication for 1 days 0 hr 0 min 0 sec
02.11.2008 11:01:21|Milkyway@home|Reason: 4 consecutive failures fetching scheduler list

Tried detaching and reattaching - to no avail.
This happened on one of my comps, while my second comp ( same model )
works just fine.

Any hints on this problem ( due to new app ?? ), or will it work out by itself ???

Thanks for any info in advance !!!

All the best

Kurt



ID: 5806 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Benjamin Rietveld

Send message
Joined: 26 Sep 08
Posts: 11
Credit: 17,597
RAC: 0
Message 5808 - Posted: 2 Nov 2008, 10:54:21 UTC

That's most likely caused by the servers being overloaded
ID: 5808 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile the silver surfer
Avatar

Send message
Joined: 27 Nov 07
Posts: 39
Credit: 1,207,109
RAC: 0
Message 5810 - Posted: 2 Nov 2008, 12:23:04 UTC - in response to Message 5808.  

That's most likely caused by the servers being overloaded


Thanks for Your answer....Then I`ll just sit it out !!!

Kurt


ID: 5810 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 5811 - Posted: 2 Nov 2008, 13:23:21 UTC

I have been keeping an eye on the server status to see if any new results ready to send are available. None have been there for the last 24 - 36 hours, and the results in progress numbers have dropped from over 24K to the current 12,644 (as of 2 Nov 2008 13:16:09 UTC).

This is the reason I made my comment 2 posts below this one.

As there is no WUs to crunch we have no choice but to sit and wait.

I am using the time to redirect my rigs to other distributed projects (mainly Einstein)
ID: 5811 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile banditwolf
Avatar

Send message
Joined: 12 Nov 07
Posts: 2425
Credit: 524,164
RAC: 0
Message 5813 - Posted: 2 Nov 2008, 13:29:30 UTC - in response to Message 5793.  

I don't think many will appear untill the next group is added. The total has gone from 36000 to 17700 wu's.


Actually my guess now is... No new work untill the "official new app" is realeased sometime in the next hundred years.
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.
ID: 5813 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 21 Aug 08
Posts: 625
Credit: 558,425
RAC: 0
Message 5815 - Posted: 2 Nov 2008, 16:39:03 UTC - in response to Message 5813.  
Last modified: 2 Nov 2008, 16:43:44 UTC

I don't think many will appear untill the next group is added. The total has gone from 36000 to 17700 wu's.


Actually my guess now is... No new work untill the "official new app" is realeased sometime in the next hundred years.


Well, all 6 I picked up a little bit ago were resends...

resultid 47150759

Created 2 Nov 2008 13:32:39 UTC
Sent 2 Nov 2008 13:32:43 UTC

You'll note that the creation date was a mere 3 minutes and 9 seconds after your message that I'm replying to, but the original replication had just timed out, so it's not a "new" task...
ID: 5815 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Dave Przybylo
Avatar

Send message
Joined: 5 Feb 08
Posts: 236
Credit: 49,648
RAC: 0
Message 5816 - Posted: 2 Nov 2008, 16:43:14 UTC - in response to Message 5815.  

We are paying attention. We are always paying attention. ;)
Dave Przybylo
MilkyWay@home Developer
Department of Computer Science
Rensselaer Polytechnic Institute
ID: 5816 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 21 Aug 08
Posts: 625
Credit: 558,425
RAC: 0
Message 5817 - Posted: 2 Nov 2008, 16:44:41 UTC - in response to Message 5816.  

We are paying attention. We are always paying attention. ;)


I'm too poor to pay attention...
ID: 5817 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
James Nelson
Avatar

Send message
Joined: 23 Nov 07
Posts: 23
Credit: 1,181,270
RAC: 0
Message 5818 - Posted: 2 Nov 2008, 17:16:30 UTC - in response to Message 5815.  



any that I've gotten last two days have been re sends and by looking at the other results from those computers it looks like they just suspended MW and let them time out.
why not abort them or detach ???
seems some are boycotting the project.

ID: 5818 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 5819 - Posted: 2 Nov 2008, 17:26:02 UTC - in response to Message 5816.  

We are paying attention. We are always paying attention. ;)


Now that is fantastic, considering Dave came in yesterday to sort the servers and it is still the week end (Sunday)

On another subject -

Prior to the Admins releasing the new faster MW client, they will need to build a reserve of WUs and take in to account the speed the WUs were mopped up by those that had installed the Milksop fast client.

As all crunchers update to the new official faster MW client, then the Admins can adjust/reduce the credit given once this is known and tested. That way they can get some cross project conformance and prevent Dr Davis A hounding them.
ID: 5819 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 5849 - Posted: 3 Nov 2008, 10:40:42 UTC
Last modified: 3 Nov 2008, 10:41:50 UTC

I have been keeping an inconsistent eye on the Server Status - results ready to send - to see if they are keeping up with the WU demand from those of us using Milksop's fast client.

Since Dave brought the servers back on line, yesterday afternoon/evening, for new work, these have coped well so far. There is not a lot of "new work" buffer present, but it is consistent at about 350 to 550 WU available (at least whenever I look).

I am aware Dave has plans to bring the servers down for some tweaking and/or adjustments prior to the release of the official MW version of the fast client. Possibly with more of slightly improved science.

I am not sure whether Dave is watching the new work server demand to see how it keeps up, as a measure of the demand when the new MW client is released.

Subject to the Admin/Dave's plans, and client release timing, it may be appropriate to let a couple of days (48 hours) pass to see how the servers cope?.
ID: 5849 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Dave Przybylo
Avatar

Send message
Joined: 5 Feb 08
Posts: 236
Credit: 49,648
RAC: 0
Message 5858 - Posted: 3 Nov 2008, 15:51:01 UTC - in response to Message 5849.  

Currently everything is run off of one server. This may have to change though since generating the work now takes up an entire CPU (out of 2) of the system consistently.
Dave Przybylo
MilkyWay@home Developer
Department of Computer Science
Rensselaer Polytechnic Institute
ID: 5858 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 21 Aug 08
Posts: 625
Credit: 558,425
RAC: 0
Message 5860 - Posted: 3 Nov 2008, 16:43:31 UTC - in response to Message 5858.  

Currently everything is run off of one server. This may have to change though since generating the work now takes up an entire CPU (out of 2) of the system consistently.


Longer running work was mentioned, but my system barely registered a difference between the gs_372 and gs_620 tasks... Truly longer running work would lighten the load...
ID: 5860 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
John Clark

Send message
Joined: 4 Oct 08
Posts: 1734
Credit: 64,228,409
RAC: 0
Message 5862 - Posted: 3 Nov 2008, 17:57:07 UTC - in response to Message 5860.  
Last modified: 3 Nov 2008, 17:57:56 UTC

Currently everything is run off of one server. This may have to change though since generating the work now takes up an entire CPU (out of 2) of the system consistently.


Longer running work was mentioned, but my system barely registered a difference between the gs_372 and gs_620 tasks... Truly longer running work would lighten the load...



You are probaby correct, Brian. But Dave's problem will be made worse when the MW fast stock client is released, hence the need for another server (a dedicated "work generation one").

As regards the longer WU, it needs to take in to account any additional science the project can usefully gain as a result.
ID: 5862 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Annika Kremer

Send message
Joined: 30 Oct 08
Posts: 32
Credit: 60,528
RAC: 0
Message 5866 - Posted: 3 Nov 2008, 20:36:09 UTC

Looks like our boxes were too much for the server to handle again. Not getting new work and the server status page shows only 38 WUs "ready to send" (which would last less than two hours on my box, so of course there's no way that's enough for even a few crunchers demanding work)...
ID: 5866 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Chris S
Avatar

Send message
Joined: 20 Sep 08
Posts: 1391
Credit: 203,563,566
RAC: 0
Message 5871 - Posted: 3 Nov 2008, 22:08:54 UTC

Message from server : server error : cant attach shared memory

started 21:57 UK time
Don't drink water, that's the stuff that rusts pipes
ID: 5871 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Annika Kremer

Send message
Joined: 30 Oct 08
Posts: 32
Credit: 60,528
RAC: 0
Message 5872 - Posted: 3 Nov 2008, 22:09:21 UTC

Looks like the feeder is not running.
ID: 5872 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
BarryAZ

Send message
Joined: 1 Sep 08
Posts: 520
Credit: 302,524,931
RAC: 15
Message 5874 - Posted: 3 Nov 2008, 22:35:12 UTC - in response to Message 5871.  

Same problem here

Message from server : server error : cant attach shared memory

started 21:57 UK time


ID: 5874 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
barsanuphe

Send message
Joined: 19 Oct 08
Posts: 19
Credit: 1,463,876
RAC: 0
Message 5877 - Posted: 3 Nov 2008, 23:17:23 UTC

things are back to normal, except credit granted is half of what it was an hour ago.
ID: 5877 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · Next

Message boards : Number crunching : Server Problems? - currently the U/L one

©2024 Astroinformatics Group