A thought on the server outage

Message boards : Number crunching : A thought on the server outage
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile Geek@Play
Volunteer tester
Avatar

Send message
Joined: 31 Jul 01
Posts: 2467
Credit: 86,146,931
RAC: 0
United States
Message 60383 - Posted: 4 Jan 2005, 17:16:33 UTC

It occurs to me that a lot of client computers are not maintaing a proper cache level. If you cannot withstand a 12 hour outage the cache level is probably set too low. Going into panic mode and punching the update button continuously does not help. Also performing a detach and re-attatch to the project is definately detrimental to the overall operation. A chache level of 2 days is what I consider the minimum level. At that setting you could probably keep crunching over a weekend outage.

OK - Flame on, I'm ready!
Boinc....Boinc....Boinc....Boinc....
ID: 60383 · Report as offensive
STE\/E
Volunteer tester

Send message
Joined: 29 Mar 03
Posts: 1137
Credit: 5,334,063
RAC: 0
United States
Message 60402 - Posted: 4 Jan 2005, 17:31:58 UTC

I agree with you Geek@Play, I usually have 4-5 days of Cache available on all my Computers.
ID: 60402 · Report as offensive
Scott Brown

Send message
Joined: 5 Sep 00
Posts: 110
Credit: 59,739
RAC: 0
United States
Message 60404 - Posted: 4 Jan 2005, 17:36:34 UTC

That's fine if your machines are on 24/7, but lower caches seem to work best with boxes that are on only part-time (e.g., during business hours, etc.)--especially if they are multi-project hosts. Also, in my case for example, I am going out of town for two weeks, so setting the cache low was necessary if I didn't want to have a bunch of late work units.
ID: 60404 · Report as offensive
.
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 410
Credit: 16,559
RAC: 0
Message 60410 - Posted: 4 Jan 2005, 17:44:20 UTC - in response to Message 60383.  
Last modified: 4 Jan 2005, 17:45:16 UTC

> It occurs to me that a lot of client computers are not maintaing a proper
> cache level. If you cannot withstand a 12 hour outage the cache level is
> probably set too low. Going into panic mode and punching the update button
> continuously does not help. Also performing a detach and re-attatch to the
> project is definately detrimental to the overall operation. A chache level of
> 2 days is what I consider the minimum level. At that setting you could
> probably keep crunching over a weekend outage.
>
> OK - Flame on, I'm ready!
>

Why would we flame you? :-)

I have my cash level set as low as 0.1 day, so I only have max. 2 WU's at the time, one in progress and one in waiting. And this suits me fine, as I won't be a bottleneck for credit granting.

And a outage can't make my day miserable, because so what? There are WU's enough for everybody, and if somebody can't live without their box crunching something, there are other projects to join! Just my oppinion!
ID: 60410 · Report as offensive
STE\/E
Volunteer tester

Send message
Joined: 29 Mar 03
Posts: 1137
Credit: 5,334,063
RAC: 0
United States
Message 60413 - Posted: 4 Jan 2005, 17:53:45 UTC

That's fine if your machines are on 24/7
=========

All my Computers run 24/7 365 days of the year .... :)
ID: 60413 · Report as offensive
1mp0£173
Volunteer tester

Send message
Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
United States
Message 60900 - Posted: 5 Jan 2005, 18:04:15 UTC - in response to Message 60404.  

> That's fine if your machines are on 24/7, but lower caches seem to work best
> with boxes that are on only part-time (e.g., during business hours,
> etc.)--especially if they are multi-project hosts. Also, in my case for
> example, I am going out of town for two weeks, so setting the cache low was
> necessary if I didn't want to have a bunch of late work units.

Hmmm....

I run part-time. Every computer that can be turned off at the end of my (incredibly long) work day is turned off. Servers stay on, but they're not running BOINC.

BOINC seems to do a good job of figuring out how much my machines are on per-day and adjusting the number of Work Units appropriately.

I agree that it's nice to flush your cache before going on vacation....
ID: 60900 · Report as offensive

Message boards : Number crunching : A thought on the server outage


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.