Welcome to MilkyWay@home

Server Trouble

Message boards : News : Server Trouble
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 22 · Next

AuthorMessage
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3339
Credit: 524,010,781
RAC: 0
Message 71913 - Posted: 10 Mar 2022, 12:13:30 UTC - in response to Message 71903.  

WOOOOOOOWWWWW!! I had no idea that you could do that and I've been doing these boards for YEARS!! Thank you Peter!!


I found it out by mistake by trying to make a duplicate message of mine just be blank. It wouldn't let me make it empty or even one space, but two worked better than I thought, it vanished. I thought I'd told you before but I must be thinking of someone else. BTW I'm up to 9 CPUs and 6 GPUs now, although not running them all 24/7 until I can afford the electricity. I've been buying up faulty GPUs and encouraging them to work. The last one usually refuses to output a display, or if it does it has blue lines down it, but it will do Milkyway or Primegrid (but nothing more complicated than that) and I can access the computer remotely so I don't need an output. I gave it a good home and it was under half price.


I don't think it was me as I think I would have remembered that.

That's cools about you finally getting more pc's and gpu's in general up and running. I'm limited right now but hopefully things will change ALOT next month when I finally get the keys to my new place, that happens the 17 of March as of right now, and then all the furniture and I can get the room for my pc's built after that. I bought some new for me parts off of a guy on Einstein and some brand new parts as well and should be able to get a couple of faster and with more cpu cores up and running in April as well. The latest stuff I got was an AMD 5800G cpu, couldn't afford the X version at the time, and 32gb of ram and an Nvidia 3060 gpu as well to go with the MB I bought off of a guy on Einstein who couldn't make 3 3080Ti gpu's work at the same time on it. It was almost brand new and a very good price as well.
ID: 71913 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3339
Credit: 524,010,781
RAC: 0
Message 71914 - Posted: 10 Mar 2022, 12:19:01 UTC - in response to Message 71912.  

No, that is up to your own settings and the initial ETA/runtime of the tasks.
Actually it happens to me all the time with many projects. Boinc is absolutely useless at working out times. I just got 2 weeks of work from LHC with my buffer set to 1 day!

To really confuse it, put a limit on how many of a certain task (say nbody) can run at once in your app config file. Boinc will senselessly keep downloading thousands of them then realising it's not allowed to run them, then get something from another project, then it will get more and more again and again.


Some of that is also the Project itself just sending tasks because it can, I had a couple of zero resource share projects send me over 400 tasks a month ago, I just aborted them and when I set the project to no new tasks and they STILL sent me another 600 tasks I aborted all of them as well!! It's not MY problem, my cache sizes are 0.5 and 0.25 respectively and like I said the project has a zero resource share set on it meaning it SHOULD only send me a single task at a time when I need it.
ID: 71914 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
David

Send message
Joined: 14 Mar 15
Posts: 2
Credit: 16,995,167
RAC: 2,385
Message 71917 - Posted: 10 Mar 2022, 15:32:12 UTC - in response to Message 71914.  

Yes,

My point was this is new. Only started happening in the last week or so, nothing do to with my settings
as nothing has changed on my end in months, Just now getting ton more worksets than usual from Milkyway
and many will be useless as they will not make deadline.

Thanks

Dave
ID: 71917 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
unixchick
Avatar

Send message
Joined: 21 Feb 22
Posts: 66
Credit: 817,008
RAC: 0
Message 71918 - Posted: 10 Mar 2022, 17:15:59 UTC - in response to Message 71917.  

I'm sorry the preferences set for MW for you aren't working properly. You seem like a seasoned pro at this, and this was working before, so I get your frustration. The only thing I can think of, and you probably already checked is to make sure that boinc is using your local preferences. As there are also preference settings on the website (https://milkyway.cs.rpi.edu/milkyway/prefs.php?subset=global)

I'm curious as I only run one project at a time. How do you have your resource share set for your different projects? I'm visiting here from WCG as they are down for 2 months. I have WCG set to 100% and MW set to 0%. So far it is behaving. I only get a new WU when one that is running is 2 minutes to finish time. Theoretically if WCG sends out any WUs then MW will stop sending WUs (I hope)
ID: 71918 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
mmonnin

Send message
Joined: 2 Oct 16
Posts: 167
Credit: 1,008,057,133
RAC: 20,257
Message 71919 - Posted: 11 Mar 2022, 0:53:37 UTC - in response to Message 71912.  

No, that is up to your own settings and the initial ETA/runtime of the tasks.
Actually it happens to me all the time with many projects. Boinc is absolutely useless at working out times. I just got 2 weeks of work from LHC with my buffer set to 1 day!

To really confuse it, put a limit on how many of a certain task (say nbody) can run at once in your app config file. Boinc will senselessly keep downloading thousands of them then realising it's not allowed to run them, then get something from another project, then it will get more and more again and again.


You described a task/user issue, not a BOINC issue.

And the 2nd thing, that is a task in progress option, not a download option. Only some projects have a download task limit setup in project preferences. It cannot be done in the client outside of the queue.
ID: 71919 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 5 Jul 11
Posts: 990
Credit: 376,143,149
RAC: 0
Message 71920 - Posted: 11 Mar 2022, 4:06:50 UTC - in response to Message 71913.  

That's cools about you finally getting more pc's and gpu's in general up and running. I'm limited right now but hopefully things will change ALOT next month when I finally get the keys to my new place, that happens the 17 of March as of right now, and then all the furniture and I can get the room for my pc's built after that. I bought some new for me parts off of a guy on Einstein and some brand new parts as well and should be able to get a couple of faster and with more cpu cores up and running in April as well. The latest stuff I got was an AMD 5800G cpu, couldn't afford the X version at the time, and 32gb of ram and an Nvidia 3060 gpu as well to go with the MB I bought off of a guy on Einstein who couldn't make 3 3080Ti gpu's work at the same time on it. It was almost brand new and a very good price as well.
Almost got to the stage of affording the electricity. About a month I think then they'll be all flat out. Cheaper tarrif sorted soon.
ID: 71920 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 5 Jul 11
Posts: 990
Credit: 376,143,149
RAC: 0
Message 71921 - Posted: 11 Mar 2022, 4:09:34 UTC - in response to Message 71918.  

I'm sorry the preferences set for MW for you aren't working properly. You seem like a seasoned pro at this, and this was working before, so I get your frustration. The only thing I can think of, and you probably already checked is to make sure that boinc is using your local preferences. As there are also preference settings on the website (https://milkyway.cs.rpi.edu/milkyway/prefs.php?subset=global)

I'm curious as I only run one project at a time. How do you have your resource share set for your different projects? I'm visiting here from WCG as they are down for 2 months. I have WCG set to 100% and MW set to 0%. So far it is behaving. I only get a new WU when one that is running is 2 minutes to finish time. Theoretically if WCG sends out any WUs then MW will stop sending WUs (I hope)
It's not %. You can set any number from 1 to a million (ish). I used to fiddle around with them, but what I do now is leave them all on 100 and turn on and off projects when I want to do them. Currently having a good go at MW, and also Sidock because some fools are boycotting it due to a couple of Russian scientists working there. Yeah those well known biologists with guns.
ID: 71921 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 5 Jul 11
Posts: 990
Credit: 376,143,149
RAC: 0
Message 71922 - Posted: 11 Mar 2022, 4:16:06 UTC - in response to Message 71919.  

You described a task/user issue, not a BOINC issue.

And the 2nd thing, that is a task in progress option, not a download option. Only some projects have a download task limit setup in project preferences. It cannot be done in the client outside of the queue.
Somebody deleted my post, can't be bothered typing all that again. I write something against Boinc and it gets removed.
ID: 71922 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile PecosRiverM

Send message
Joined: 25 Aug 17
Posts: 12
Credit: 1,248,922,375
RAC: 145,125
Message 71926 - Posted: 11 Mar 2022, 17:39:24 UTC - in response to Message 71922.  

I can't even get mine u/l'd. they deadline in 14 hours and have been trying the last 5 hours.

3/11/2022 11:35:30 AM | Milkyway@Home | <![CDATA[update requested by user]]>
3/11/2022 11:35:41 AM | | <![CDATA[Project communication failed: attempting access to reference site]]>
3/11/2022 11:35:43 AM | | <![CDATA[Internet access OK - project servers may be temporarily down.]]>
ID: 71926 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile PecosRiverM

Send message
Joined: 25 Aug 17
Posts: 12
Credit: 1,248,922,375
RAC: 145,125
Message 71933 - Posted: 12 Mar 2022, 2:24:47 UTC - in response to Message 71926.  

Thank You (whoever kicked the Server).

back to u/l
ID: 71933 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Dave Studdert
Avatar

Send message
Joined: 26 Mar 09
Posts: 2
Credit: 22,654,013
RAC: 9,045
Message 71948 - Posted: 14 Mar 2022, 13:36:48 UTC

This explains the lack of work for the nvidia cards. ati card still has a decent queue of work. Current Validation pending (4245)
ID: 71948 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom George
Avatar

Send message
Joined: 29 Dec 21
Posts: 7
Credit: 8,995,805
RAC: 0
Message 71950 - Posted: 14 Mar 2022, 14:57:33 UTC

How goes the server battle Tom? Anything we (as a community) can help with?
ID: 71950 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 5 Jul 11
Posts: 990
Credit: 376,143,149
RAC: 0
Message 71951 - Posted: 14 Mar 2022, 15:29:43 UTC - in response to Message 71950.  

How goes the server battle Tom? Anything we (as a community) can help with?
Cash I would think, but he doesn't seem to want charity. Many Boinc projects accept donations.
ID: 71951 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom Donlon
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 10 Apr 19
Posts: 408
Credit: 120,203,200
RAC: 0
Message 71952 - Posted: 14 Mar 2022, 17:58:54 UTC

Apologies for the delay in my updates. I was off the last week, but now I'm back.

We do accept donations! The page for that is located here: https://securelb.imodules.com/s/1225/giving/index.aspx?sid=1225&gid=1&pgid=3676. We used to accept gridcoin donations from our users, but that generated several problems with RPI/the IRS, so we can't do that anymore.

As for the server battle, I'm just trying to keep things afloat until we can get this drive back. I've run the transitioner script to remove that enormous backlog, and I've flushed the DB of any stuck jobs. I also restarted all the server's major processed. Hopefully we should see the number of available jobs come back up, because right now we're out of them.

I can always restart the server soon if we don't see any changes - usually that helps with the memory load for a couple days.
ID: 71952 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
FurryGuy

Send message
Joined: 1 Aug 11
Posts: 10
Credit: 51,374,490
RAC: 0
Message 71953 - Posted: 14 Mar 2022, 22:24:33 UTC - in response to Message 71952.  

Thank you for keeping us "in the loop."

It is frustrating to have the project's servers down, more so when we crunchers don't know what is happening.
ID: 71953 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom Donlon
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 10 Apr 19
Posts: 408
Credit: 120,203,200
RAC: 0
Message 71954 - Posted: 15 Mar 2022, 1:01:41 UTC

I ended up restarting the server and doing some more behind the scenes DB maintenance. Hopefully some jobs start going out, because the project status page shows 0 separation and 0 nbody tasks right now.
ID: 71954 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Speedy51

Send message
Joined: 12 Jun 10
Posts: 57
Credit: 6,171,817
RAC: 8
Message 71955 - Posted: 15 Mar 2022, 2:17:57 UTC
Last modified: 15 Mar 2022, 3:15:14 UTC

This is just an idea. Tom have you thought about turning work creation off allowing the resends to be created if possible until the administrator and the work pending validation that has enough copies returned to validate has caught up to allow some space to be cleared off with the disc? This may also help speed up the work creation rate.
I have had 6 _2 tasks all of these have returned waiting validation. I am aware this will happen when the server catches up
Thanks for all the work you are doing
ID: 71955 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Cameron

Send message
Joined: 16 Dec 07
Posts: 37
Credit: 25,673,575
RAC: 5,663
Message 71956 - Posted: 15 Mar 2022, 2:57:18 UTC

Just got a Single MW Seperation task when I reurned all my completed work.
I'll try again later when the backlog has had a chance to clear and the gererator some time to create WUs
ID: 71956 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 5 Jul 11
Posts: 990
Credit: 376,143,149
RAC: 0
Message 71957 - Posted: 15 Mar 2022, 4:07:27 UTC - in response to Message 71952.  

We do accept donations! The page for that is located here: https://securelb.imodules.com/s/1225/giving/index.aspx?sid=1225&gid=1&pgid=3676.
You should have that on the home page like some other projects do (eg. PrimeGrid, Sidock), I doubt many know about it.

We used to accept gridcoin donations from our users, but that generated several problems with RPI/the IRS
Typical.
ID: 71957 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Tom George
Avatar

Send message
Joined: 29 Dec 21
Posts: 7
Credit: 8,995,805
RAC: 0
Message 71958 - Posted: 15 Mar 2022, 13:31:10 UTC

Looks like it's back up and working again!
ID: 71958 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 22 · Next

Message boards : News : Server Trouble

©2024 Astroinformatics Group