Limiting BRP4 to CUDA machines

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 417526914
RAC: 33416

RE: SETI has now started

Quote:
SETI has now started generating work again, so I expect the Einstein demand will decrease over the course of the weekend as users switch back - that will allow our WUGs to have a better go at maintaining a supply.

My machines still suffer lack of BRP4 CUDA tasks. It is hard to receive CUDA tasks. I have to force multiple BOINC requests to receive tasks at least for a day. SETI doesn't help to resolve this issue. So, we are hungry to crunch some new numbers.

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2139
Credit: 2753018280
RAC: 1371263

RE: RE: SETI has now

Quote:
Quote:
SETI has now started generating work again, so I expect the Einstein demand will decrease over the course of the weekend as users switch back - that will allow our WUGs to have a better go at maintaining a supply.

My machines still suffer lack of BRP4 CUDA tasks. It is hard to receive CUDA tasks. I have to force multiple BOINC requests to receive tasks at least for a day. SETI doesn't help to resolve this issue. So, we are hungry to crunch some new numbers.


SETI's work generation phase lasted for less than 36 hours (this time) - they ran out of raw data on Friday evening (UTC), and have been unable to supply work since then. The added pressure on our servers continues.

Filipe
Filipe
Joined: 10 Mar 05
Posts: 175
Credit: 365552287
RAC: 45129

Seti is producing new work.

Seti is producing new work.

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 417526914
RAC: 33416

RE: RE: RE: SETI has

Quote:
Quote:
Quote:
SETI has now started generating work again, so I expect the Einstein demand will decrease over the course of the weekend as users switch back - that will allow our WUGs to have a better go at maintaining a supply.

My machines still suffer lack of BRP4 CUDA tasks. It is hard to receive CUDA tasks. I have to force multiple BOINC requests to receive tasks at least for a day. SETI doesn't help to resolve this issue. So, we are hungry to crunch some new numbers.

SETI's work generation phase lasted for less than 36 hours (this time) - they ran out of raw data on Friday evening (UTC), and have been unable to supply work since then. The added pressure on our servers continues.

It is rather strange. There are many other CUDA projects, but it seems SETI users prefer to run E@H as a reserve project whereas other projects give more points per hour then yet they do not suffer of SETI machines.

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 417526914
RAC: 33416

Something happend. Queue

Something happend. Queue begins to grow.
I wonder why BRP4 work generators ## 0,7,8,9 and 18 are disabled all the time? Are they for some special tasks?

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933581
RAC: 16281

RE: Something happend.

Quote:
Something happend. Queue begins to grow.

I figure some GPUs got work from SETI and thus don't ask work from E@H.

Quote:
I wonder why BRP4 work generators ## 0,7,8,9 and 18 are disabled all the time? Are they for some special tasks?

These are just configuration slots.

WUG #1-6 run on a second machine that doesn't have enough resources (disk space, cores) to run more than these 6 instances.

WUGs #10-19 run on the original machine (einstein-wug). Enabling more than the currently running 9 instances makes no sense there. Every additional instance running on that machine slows down the runtime per beam of all WUGs by roughly an hour. I.e. although the system load would be higher, the number of WUs produced per hour would stay the same. In addition, mindlessly overloading this system would be bad, since the FGRP daemons and the other BRP daemons (validator, assimilator etc.) are running on this machine, too.

BM

BM

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 417526914
RAC: 33416

It gets worse. BTW, it is the

It gets worse. BTW, it is the end of "the next week". Any news about configuration changes?

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244933581
RAC: 16281

RE: It gets worse. BTW, it

Quote:
It gets worse. BTW, it is the end of "the next week". Any news about configuration changes?

This is not about changing configuration. It means designing and implementing a new scheme of workunit generation. Oliver and Benjamin are working on it.

Additionally last night a network cable broke that connected the 'relief WUG machine'. All BRP4 workunit generation was stopped for ~10h.

BM

BM

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 417526914
RAC: 33416

But why just not to install

But why just not to install an additional relief machine and get rid of the problem instead of implementing a new mechanism in a hurry? New scheme can be then implemented whith the start of the next run. Or is it a heat wave coming from the additional rig and it do not allow to work? ;)

robertmiles
robertmiles
Joined: 8 Oct 09
Posts: 127
Credit: 20996335
RAC: 79119

Have you donated enough to

Have you donated enough to pay for the additional relief machine? They do not have an unlimited budget, even if there's no heat problem.

Or would you prefer to donate only enough to buy a new cable for connecting to the relief machine they already have?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.