| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Official Messages
Forum: News
Thread: 2022-08-08 update (new workunits, FAQ thread soon) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 45
|
|
| Author |
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 1934 Status: Offline Project Badges: 
|
This is from Cyclops on the 19th. I think the main problem is that there isn't a new update that is dated every few days. People would know where the latest update was, and old issue threads could fade away on the forum. That makes little sense. Those networking issues and any database/WCG stats issues, that are two different ball games, that should handled by different people. Unless the WCG administration is handled by one single person, who just happened to be on summer break the last two months.."While we continue to work at a lower capacity, we have identified an additional path to increasing our capacity by additional 10-20% as we await the fix by co-locating select services on our QA and build servers. As we configure the system to handle peak loads and prepare to deploy the resources that are blocked by the networking issue at the data center, we will continue to update you. With this issue on a path to resolution, we should now be able to address some of the concerns over statistics, credit, streaks, and database dumps which we have already begun to investigate. Ralf PS: And someone must have already "played" with this, as while I was "playing" with various versions of the history.csv file from various points in time today, it seems that the credits in that file are no longer those WCG credits of old, which used to be 7x that of the BOINC credits. Which needs to be considered or that stat on the Overview page is getting totally trashed...  |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 1934 Status: Offline Project Badges:
|
It's just that at this stage of the restart of WCG, I do not find it important at all. I can follow my crunching (credits), on the external stat sites, and I'm fully content with that for now. In due time, the WCG team will fix the internal stats (history, teams, and whatever else). I'm not bothered at all with that problem for now. At this stage of the restart, there simply might not be any data around anymore to even try and fix that issue. As the discrepancy between the two text files that I mentioned so far shows. And there is no inkling on Krembil's side that this data has been in fact retained.And two months (after a previous 5+4 months), this is for me hard to believe... I'd like to be wrong, but until I am proved wrong, my statements stay... Ralf |
||
|
|
debs3759
Cruncher Joined: Jun 10, 2021 Post Count: 20 Status: Offline Project Badges: 
|
I'm not sure which thread this should go in, so will post it here and hope it is right
I've noticed that mcm1.dataset-sarc1.txt keeps getting downloaded. It's over 100MB, and I can't see why it needs downloading on a regular basis, even when already on my system. |
||
|
|
Jean-David Beyer
Senior Cruncher USA Joined: Oct 2, 2007 Post Count: 335 Status: Offline Project Badges: 
|
I've noticed that mcm1.dataset-sarc1.txt keeps getting downloaded. It's over 100MB, and I can't see why it needs downloading on a regular basis, even when already on my system. I notice that. once the thing runs, it copies it to whatever slot is appropriate. I have three instances of the work unit running, and only one mcm1.dataset-sarc1.txt file up in the /var/lib/boinc/projects/www.worldcommunitygrid.org directory. The Boinc-server certainly cannot be bothered to keep track of every file downloaded to every client, so I would hope the Boinc-client could check that for each file coming down. Of course for that to be possible, each such file would need to send identifying information so the client could tell the difference between a duplicate and an update. I do not imagine the Boinc-client actually works this way.  |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 885 Status: Offline Project Badges:
|
I've noticed that mcm1.dataset-sarc1.txt keeps getting downloaded. It's over 100MB, and I can't see why it needs downloading on a regular basis, even when already on my system. When you run out of work for a particular [sub-]project, if the next contact with the server doesn't fetch more work for that application it might result in the client tidying up some files that don't disappear when the associated task reports in. (I think the server initiates the "out of work" clean-up; it does get a comprehensive list of files on the client for the [main] project...) So, if you run out of MCM1 work and the next contact with the server doesn't snag you some more, it'll tell the client to "tidy up" and that big file disappears :-( Then, the next time you do get some MCM1 work the server knows that file isn't on the client and it sends it again... I suspect it's marked as removable because of its size (which might be quite a big chunk out of a small system's disk space!); the executable files and any graphics don't get removed unless there's a project reset or removal... Cheers - Al. |
||
|
|
|