Welcome to MilkyWay@home

Down for maintenance?


Advanced search

Message boards : Number crunching : Down for maintenance?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · Next

AuthorMessage
SkyeHunter

Send message
Joined: 6 Mar 09
Posts: 41
Credit: 38,856,291
RAC: 0
30 million credit badge10 year member badge
Message 37204 - Posted: 11 Mar 2010, 11:05:26 UTC

I noticed that, prior to the outages, I was crunching short running WU's only...

I only have a view on the half hour prior to the outage of 3h40 and at that time I was in bed, so no live view...

And no, I don't have a script deleting longer, normal WU's.
ID: 37204 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
UBT - JohnR

Send message
Joined: 10 Mar 08
Posts: 7
Credit: 60,169,291
RAC: 0
50 million credit badge10 year member badge
Message 37206 - Posted: 11 Mar 2010, 12:46:22 UTC - in response to Message 37204.  
Last modified: 11 Mar 2010, 13:06:05 UTC

I noticed that all the work was short WU's, and that put a load on the server that would have been like a denial of service attack.

If we were allowed to download more work, say 24 WU's per core and then denied any more work for the next 5 minutes instead of 1 minute, the servers would be spared all the fast machines asking for work every minute. I would guess that the servers can give out 5 WU's per request almost as easily as 1.

That should take a great load off the servers. I have my cache set to 0.02 days so I don't ask for work when there is no completed work.

The trouble then is that the CPU's can be run out of work on other projects.
ID: 37206 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
ProfileThe Gas Giant
Avatar

Send message
Joined: 24 Dec 07
Posts: 1947
Credit: 240,884,648
RAC: 0
200 million credit badge10 year member badge
Message 37222 - Posted: 11 Mar 2010, 19:50:02 UTC - in response to Message 37206.  

...

If we were allowed to download more work, say 24 WU's per core and then denied any more work for the next 5 minutes instead of 1 minute, the servers would be spared all the fast machines asking for work every minute. I would guess that the servers can give out 5 WU's per request almost as easily as 1.

That should take a great load off the servers. I have my cache set to 0.02 days so I don't ask for work when there is no completed work.

The trouble then is that the CPU's can be run out of work on other projects.

From a users perspective it sounds like a good idea, but the project admins have previously told us that if there are too many wu's in the DB it slows down the accessing of the DB leading to crashes which causes us all problems. So until the server gets a boost (if ever) we are stuck with what we have.
ID: 37222 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
loeakaodas

Send message
Joined: 2 Jan 09
Posts: 34
Credit: 93,631,891
RAC: 0
50 million credit badge10 year member badge
Message 37462 - Posted: 17 Mar 2010, 22:54:59 UTC

Looks like it's down again, for at least the last 5 hours.
ID: 37462 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile[B@H] Ray

Send message
Joined: 27 Dec 07
Posts: 35
Credit: 1,432,926
RAC: 0
1 million credit badge10 year member badge
Message 37463 - Posted: 17 Mar 2010, 23:16:08 UTC

I thought it was just me getting that.....
ID: 37463 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilebanditwolf
Avatar

Send message
Joined: 12 Nov 07
Posts: 2425
Credit: 524,164
RAC: 0
500 thousand credit badge10 year member badge
Message 37464 - Posted: 18 Mar 2010, 0:10:05 UTC

I am too...
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.
ID: 37464 · 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
50 million credit badge10 year member badge
Message 37468 - Posted: 18 Mar 2010, 0:35:58 UTC

This loss of work has been going on for many hours now, and [b]the server status makes it look like it will continue.[b]

It is time for sleep so I will run Collatz until I am awake and can revert, assuming the servers are back
Go away, I was asleep


ID: 37468 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
ProfileThe Gas Giant
Avatar

Send message
Joined: 24 Dec 07
Posts: 1947
Credit: 240,884,648
RAC: 0
200 million credit badge10 year member badge
Message 37472 - Posted: 18 Mar 2010, 1:40:44 UTC

Damn! And just when my RAC had hit record levels!

I wonder if Collatz will fall over with the influx of users.....
ID: 37472 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilewdsmia
Avatar

Send message
Joined: 20 Nov 07
Posts: 5
Credit: 281,219,758
RAC: 0
200 million credit badge10 year member badge
Message 37476 - Posted: 18 Mar 2010, 2:18:00 UTC


Damn! And just when my RAC had hit record levels!


Ohh, you poor fella! ;p
ID: 37476 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profilearkayn
Avatar

Send message
Joined: 14 Feb 09
Posts: 999
Credit: 74,932,619
RAC: 0
50 million credit badge10 year member badge
Message 37480 - Posted: 18 Mar 2010, 3:32:51 UTC

This time I was closing in on 100,000 RAC.

New video cards tend to do that.
ID: 37480 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Misfit
Avatar

Send message
Joined: 27 Aug 07
Posts: 915
Credit: 1,503,319
RAC: 0
1 million credit badge10 year member badge
Message 37482 - Posted: 18 Mar 2010, 4:39:05 UTC - in response to Message 35654.  

Database has been in down for maintenance for a few hours???

I'm trying to figure out what exactly happened. Hopefully I'll have more info available later when I hear from labstaff.

Exiled to the mines of Cosmo.
me@rescam.org
ID: 37482 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
BarryAZ

Send message
Joined: 1 Sep 08
Posts: 519
Credit: 283,765,560
RAC: 5,748
200 million credit badge10 year member badgeextraordinary contributions badge
Message 37483 - Posted: 18 Mar 2010, 5:36:07 UTC - in response to Message 37482.  

Here we go again with an outage of indefinite length with an information outage as well. And the load shift over to Collatz has the same effect...
ID: 37483 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
ProfileBerserk_Tux
Avatar

Send message
Joined: 2 Jan 08
Posts: 79
Credit: 365,471,675
RAC: 0
300 million credit badge10 year member badge
Message 37487 - Posted: 18 Mar 2010, 11:41:16 UTC - in response to Message 37483.  

Nothing from Travis and the prosjekt is still down.
ID: 37487 · 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
50 million credit badge10 year member badge
Message 37489 - Posted: 18 Mar 2010, 12:09:36 UTC

Getting on for 24 hours now
Go away, I was asleep


ID: 37489 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
STE\/E

Send message
Joined: 29 Aug 07
Posts: 486
Credit: 573,906,610
RAC: 0
500 million credit badge10 year member badge
Message 37490 - Posted: 18 Mar 2010, 12:12:44 UTC - in response to Message 37483.  
Last modified: 18 Mar 2010, 12:13:06 UTC

Here we go again with an outage of indefinite length with an information outage as well. And the load shift over to Collatz has the same effect...


Collatz seems slow but finished work is getting Uploaded and new work Downloaded on my Box's, very slowly though ...
STE\/E
ID: 37490 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Skysnake

Send message
Joined: 31 Oct 09
Posts: 20
Credit: 12,074,198
RAC: 0
10 million credit badge10 year member badge
Message 37491 - Posted: 18 Mar 2010, 12:28:21 UTC - in response to Message 37490.  
Last modified: 18 Mar 2010, 12:30:36 UTC

Here we go again with an outage of indefinite length with an information outage as well. And the load shift over to Collatz has the same effect...


Collatz seems slow but finished work is getting Uploaded and new work Downloaded on my Box's, very slowly though ...


Yes, Collatz hp is down nearly, but you get work and can send work. So the important thinks work ;) I have 20 WU´s like allways.

But i also think, that there should be a news with updates on the mainpage!

90% off my GPU time should be used for MW, so i often just stop Collatz to fetch new work (YES i know not the best way ;) ). So i waste yesterday about 2 hours of GPUtime. Idle sucks :/

EDIT: Ok Collatz is down atm ;( no report or new work ;(
ID: 37491 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
ProfileVincent JG
Avatar

Send message
Joined: 25 Jan 10
Posts: 127
Credit: 118,190,343
RAC: 0
100 million credit badge10 year member badge
Message 37493 - Posted: 18 Mar 2010, 14:21:05 UTC

Lol, maybe today wasn't the best launch date for my new cruncher. Collatz seems down too, stuck at uploading/downloading
"Government big enough to supply everything you need is big enough to take everything you have."
-Thomas Jefferson
American Thinker
ID: 37493 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
BarryAZ

Send message
Joined: 1 Sep 08
Posts: 519
Credit: 283,765,560
RAC: 5,748
200 million credit badge10 year member badgeextraordinary contributions badge
Message 37497 - Posted: 18 Mar 2010, 15:12:33 UTC - in response to Message 37493.  

This seems quite predictable. MW goes offline, unannounced, unexplained and with no news updates as to what is going on or when (if) it will be resolved. That then spurs a flock of 90/10 MW/Collatz folks to go 100% Collatz.

Collatz, inundated by the tsunami of temporary MW refugees, gets swamped and encounters surge load problems that it as a one person unfunded shop can't readily handle.

The interesting thing is that typically not only does Collatz seem to recover from the surge before MW is back on line, Collatz provides an explanation as to what happened for them long before we get blessed with an explanation over here.

Personally, MW became my number 2 project some time ago, thanks to Collatz support for single precision GPU AND the proactive communication available over there.

I get the impression that if MW loaned Collatz some of its resources that would project planning/operation and communications training that would have value for them here in MW land.
ID: 37497 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
55degrees

Send message
Joined: 8 Sep 09
Posts: 62
Credit: 61,330,584
RAC: 0
50 million credit badge10 year member badge
Message 37500 - Posted: 18 Mar 2010, 15:15:28 UTC - in response to Message 37491.  

"90% off my GPU time should be used for MW, so i often just stop Collatz to fetch new work (YES i know not the best way ;) ). So i waste yesterday about 2 hours of GPUtime. Idle sucks :/"

skysnake, please excuse the out of place post: maybe I misunderstand, but how are you setting 90% gpu for mw having collatz as another share? I have had no success running collatz with mw in that collatz takes over no matter how I set the sharing. I read that it might be a boinc problem, but whatever it is collatz dictates on my machine if running. that is not happening to you?


thanks.
ID: 37500 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Skysnake

Send message
Joined: 31 Oct 09
Posts: 20
Credit: 12,074,198
RAC: 0
10 million credit badge10 year member badge
Message 37501 - Posted: 18 Mar 2010, 16:56:25 UTC - in response to Message 37500.  

"90% off my GPU time should be used for MW, so i often just stop Collatz to fetch new work (YES i know not the best way ;) ). So i waste yesterday about 2 hours of GPUtime. Idle sucks :/"

skysnake, please excuse the out of place post: maybe I misunderstand, but how are you setting 90% gpu for mw having collatz as another share? I have had no success running collatz with mw in that collatz takes over no matter how I set the sharing. I read that it might be a boinc problem, but whatever it is collatz dictates on my machine if running. that is not happening to you?


thanks.


NP ;)

The problem is very simpel. A MW WU takes about 1:40 and a Collatz WU about 5:40
MW allows 12 WU´s max. Collatz about 20 or more. So i spend allways more time for collatz, because if a WU is uploaded, a new WU is downloaded. Perhaps just chance the preferences should help, but i haven´t try it yet.
ID: 37501 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · 4 · 5 · Next

Message boards : Number crunching : Down for maintenance?

©2020 Astroinformatics Group