mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   GPU to 72 (https://www.mersenneforum.org/forumdisplay.php?f=95)
-   -   Assignments (https://www.mersenneforum.org/showthread.php?t=23276)

storm5510 2018-04-22 00:29

Assignments
 
2 Attachment(s)
I have a bit of a conundrum it seems. The two dialogs below do not seem to agree. The first image has a red stripe indicating just over 500 GHz-d/days in its queue.The second is set to 1050 GHz-d/days to fetch. What am I missing? :question:

chalsall 2018-04-22 14:56

[QUOTE=storm5510;485909]What am I missing? :question:[/QUOTE]

First of all, thanks for joining the GPU72 crew! :smile:

I'm afraid I can't speak to your question; I don't run MISFIT myself (my two little GPUs are running under Linux). Perhaps Mark or Wayne can give you some clarity, but at the moment you've got 2.5 days of work queued up.

LaurV 2018-04-22 16:02

1 Attachment(s)
The process is that you add (fetch) work from the servers which is kept in a local file. From the local file, the work is sucked (drained) by all the systems/cores involved (which can be distributed in a network, on a large area). This is called assigning work, or draining. What you miss is that Misfit can coordinate the work of many systems, and all these systems can have different working preferences, speeds, etc. You have only one system, but the process is still the same, the local file into which you "fetch" work from the server is practically a buffer in case you can't connect for a while. So, you have to fetch work, AND assign work. The processes are independent - it should theoretically be "shorted" when you have a single system, but practically it is not. You fetch 1300 GHzDays every time, but you still have to look for a "third" picture (beside the two posted) where you assign this work to your system. Like the "work management" window, with the draining settings.
[ATTACH]18126[/ATTACH]

petrw1 2018-04-22 16:56

[QUOTE=storm5510;485909]I have a bit of a conundrum it seems. The two dialogs below do not seem to agree. The first image has a red stripe indicating just over 500 GHz-d/days in its queue.The second is set to 1050 GHz-d/days to fetch. What am I missing? :question:[/QUOTE]

Has it improved? It may take it a second try to get the rest of the work?

storm5510 2018-04-23 02:25

[QUOTE=chalsall]First of all, thanks for joining the GPU72 crew! :smile:[/QUOTE]

Thank you! Most kind.

The real problem was that I did not allow it to run long enough.

Since there was not much to read about how it worked, I simply experimented with the settings. It was giving me a [U]lot[/U] of assignments, at first. The staged GHz-d/day was once over 2,500. I have gradually gotten this down to about 12 hours worth, staged + active.

There have been two occasions where the command prompt console disappeared. At first, I did not realize that this meant it had stopped. I will attribute this to a Windows 10 issue. It does not seem to like anything not new and shiny.

:smile:


All times are UTC. The time now is 13:57.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.