Message boards :
Number crunching :
Estimate on How Long We're Going to be Out of CurieDock WUs :)
Message board moderation
Author | Message |
---|---|
Send message Joined: 24 Oct 20 Posts: 5 Credit: 6,335,145 RAC: 0 |
Title says it all. |
Send message Joined: 9 Oct 20 Posts: 185 Credit: 2,782,517 RAC: 142 |
We are creating new WUs in half an hour. We are sorry for inconveniences. It will take a while to adapt to the growing number of participants/computers, while server's disc capacity is limited. |
Send message Joined: 5 Dec 20 Posts: 8 Credit: 2,973,433 RAC: 4,172 |
it would help, if everybody would load just the amount of wus, which he can calculate at once + maybe the same amount for buffer. |
Send message Joined: 30 Oct 20 Posts: 57 Credit: 9,112,528 RAC: 0 |
it would help, if everybody would load just the amount of wus, which he can calculate at once + maybe the same amount for buffer. Yes. Keeping a large buffer just delays the project for everyone. I normally use the default of 0.1 + 0.5 days for work units of normal length. That works fine with this project. |
Send message Joined: 5 Dec 20 Posts: 8 Credit: 2,973,433 RAC: 4,172 |
Yes. Keeping a large buffer just delays the project for everyone. I normally use the default of 0.1 + 0.5 days for work units of normal length. I usually have 0 + 0 days and a backup project with ressource share 0, works fine |
Send message Joined: 30 Oct 20 Posts: 57 Credit: 9,112,528 RAC: 0 |
I usually have 0 + 0 days and a backup project with ressource share 0, works fine That certainly minimizes the work units to the minimum possible. But the reason for a buffer is to also reduce the load on the server so that it does not have to be accessed for each work unit. I don't know if that is an important consideration for the servers on this project or not, but that was the original idea I believe, in addition to giving you some cushion in case of a server outage of course. |
Send message Joined: 26 Oct 20 Posts: 53 Credit: 2,520,836 RAC: 0 |
I don't know if that is an important consideration for the servers on this project or not, but that was the original idea I believe, in addition to giving you some cushion in case of a server outage of course.The workunits are sent with the flag <report_immediately/>, so the project wants them back asap. When reported you'll get new tasks up to the server limit 2/thread. |
©2024 SiDock@home Team