| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: World Community Grid Moves to IBM Cloud |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 182
|
|
| Author |
|
|
deltavee
Ace Cruncher Texas Hill Country Joined: Nov 17, 2004 Post Count: 4894 Status: Offline Project Badges: 
|
2. FA@H - 2 - Bedam does not have the trickle handler running at all yet. These work units have multiple parts to them. The intermediate uploads need to complete, then trickle messages get processed. Because of the large outage, I have disabled the trickle message handler so that members had 24 hours to return those intermediate files. We are seeing uploads coming in at a consistent rate, so the surge of uploads are over. I will turn on the trickle message handler first and let it process those messages. After it has caught up, then the validator will be turned on to allow for final credit of the result. Once the validator is caught up, then the process to monitor late results will be started. Then it should be business as usual. Thanks, -Uplinger Thanks, Uplinger As one of those who chose to ride through the outage crunching FA@H - 2 exclusively, I would like to thank you for making this extra effort. |
||
|
|
wchoff
Cruncher Joined: Nov 17, 2004 Post Count: 35 Status: Offline Project Badges: 
|
I'm having similar trouble as Greatnessguru. Before the outage I set my cache size larger than usual (4 days instead of 0.3). Yesterday, once the website was back up I was able to revert it back to the old value and save. But this morning I see I am downloading new work to maintain a 4 day cache even as the 0.3 value is saved.
|
||
|
|
SekeRob
Master Cruncher Joined: Jan 7, 2013 Post Count: 2741 Status: Offline |
There's 2 spots, first spot the website device profiles and the local preferences. For new settings to take effect, they first need to transfer and during profile update transfer, the client still only knows the old setting, therefor asks for work per your 4 days. The second time, there wont be work requesting until your lower setting takes effect.
----------------------------------------Recommendation: If lowering buffer substantially, first suspend work fetch, then hit update to transfer new settings from web profile. ** Spot 2 , local, overrides web settings, until you reset them. edit: ** It goes without saying, that work fetching after the update succeeded, profile update time printed in log, needs setting to allow again. [Edit 2 times, last edit by SekeRob* at May 16, 2017 4:40:47 PM] |
||
|
|
wchoff
Cruncher Joined: Nov 17, 2004 Post Count: 35 Status: Offline Project Badges:
|
I understand all that Sek. I've never changed the local preferences, and I had a steady stream of new units, dozens of them, for about 8 hours after work started going out again. The client should have 'known' the new setting by then. It certainly was set much more quickly when I raised the cache size.
The good news is, the flow seems to have stopped. No new units for several hours now. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Good for you, you all understand that, but not everybody does, in that my smell is not perfect.
Of course saving profiles and the resumption of the various messaging streams could have led to delay, which is why the tidbit was added about the (event) log printing what date/timestamp of the last profile change is the positive feedback that the client forward on knows. Did you check those logs ;? |
||
|
|
widdershins
Veteran Cruncher Scotland Joined: Apr 30, 2007 Post Count: 677 Status: Offline Project Badges: 
|
Still getting a Server under maintenance message when trying to upload SCC units. Is this a cached DNS issue at my end, or something at WCG end still to be sorted? Been playing with the logging flags, is this line in the middle significant? Tue 16 May 2017 21:20:16 BST World Community Grid Backing off 33 min 41 sec on upload of SCC1_0000416_Lin-CSD-D_93339_0_r1452507876_0 Tue 16 May 2017 21:20:16 BST World Community Grid [fxd] starting upload, upload_offset -1 Tue 16 May 2017 21:20:16 BST World Community Grid Started upload of SCC1_0000415_Lin-CSD-D_97183_0_r112261451_0 Tue 16 May 2017 21:20:16 BST World Community Grid [file_xfer_debug] URL: https://grid.worldcommunitygrid.org/boinc/wcg_cgi/file_upload_handler Tue 16 May 2017 21:20:16 BST [poll_debug] CLIENT_STATE::poll_slow_events(): pers_file_xfers Tue 16 May 2017 21:20:16 BST [poll_debug] CLIENT_STATE::do_something(): End poll: 3 tasks active Tue 16 May 2017 21:20:16 BST [poll_debug] CLIENT_STATE::do_something(): End poll: 0 tasks active Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: About to connect() to grid.worldcommunitygrid.org port 443 (#2) Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Trying 198.20.8.241... Tue 16 May 2017 21:20:16 BST [network_status_debug] status: online Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Connected to grid.worldcommunitygrid.org (198.20.8.241) port 443 (#2) Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: successfully set certificate verify locations: Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: CAfile: ca-bundle.crt Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: CApath: /etc/ssl/certs Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: SSLv3, TLS handshake, Client hello (1): Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Unknown SSL protocol error in connection to grid.worldcommunitygrid.org:443 Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Expire cleared Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Closing connection #2 It's a couple of older machines with an old version of linux and old version of Boinc. Never bothered upgrading them as they only crunch for WCG and have bombproof reliability (both in system uptime, and returning 100% valids) the only time they've fallen over was when there is a power cut. Did you up the minimum version of SSL at the switchover? If so could you reduce it again please? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: About to connect() to grid.worldcommunitygrid.org port 443 (#2) grid.worldcommunitygrid.org should now be 169.47.63.74, so it looks as if you do have a DNS issue.Tue 16 May 2017 21:20:16 BST [http_debug] [ID#40118] Info: Trying 198.20.8.241... |
||
|
|
widdershins
Veteran Cruncher Scotland Joined: Apr 30, 2007 Post Count: 677 Status: Offline Project Badges:
|
Ah, well that helps narrow the problem down. Thanks for that, hopefully it'll be an easy fix.
|
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
I have started up the trickle message handler for Fight AIDS @ Home.
Thanks, -Uplinger |
||
|
|
keithhenry
Ace Cruncher Senile old farts of the world ....uh.....uh..... nevermind Joined: Nov 18, 2004 Post Count: 18667 Status: Offline Project Badges: 
|
Found a nit - the search index for the forums needs to be rebuilt. Can't search the forums without getting that Lucene search index error.
---------------------------------------- |
||
|
|
|