| *** manager has joined #schooltool | 04:54 | |
| *** yvl has joined #schooltool | 07:35 | |
| *** yvl has quit IRC | 07:37 | |
| *** yvl has joined #schooltool | 07:38 | |
| *** robb_nl has joined #schooltool | 13:30 | |
| *** robb_nl has quit IRC | 15:44 | |
| *** robb_nl has joined #schooltool | 16:35 | |
| *** robb_nl has quit IRC | 16:59 | |
| *** th1a has joined #schooltool | 17:43 | |
| *** robb_nl has joined #schooltool | 18:04 | |
| *** yvl has quit IRC | 18:19 | |
| *** replaceafill has joined #schooltool | 18:27 | |
| th1a | hi replaceafill. | 18:29 |
|---|---|---|
| replaceafill | hey th1a | 18:29 |
| th1a | Well, it seems like we have a few issues in play, none of which require you to do any actual coding. | 18:30 |
| th1a | So hopefully I can mostly handle the emails. | 18:30 |
| replaceafill | few? any other besides Steph's? | 18:30 |
| th1a | Well, it is multi-part. | 18:31 |
| th1a | And I should probably explain blueprints a bit. | 18:32 |
| replaceafill | ok | 18:32 |
| th1a | that is, everyone has access but only Ayesha should be on the team. | 18:33 |
| th1a | (and Fran and us) | 18:35 |
| replaceafill | ok | 18:36 |
| th1a | So we need to write an email about 32/64 bit and probably also reviewing repozo. | 18:36 |
| th1a | we/I | 18:36 |
| replaceafill | i'm not sure it's 32/64 | 18:36 |
| th1a | Ah. | 18:36 |
| th1a | What are you thinking? | 18:36 |
| replaceafill | problem with this kind of issues is that Steph is explaining things bit by bit | 18:37 |
| replaceafill | saw the "errors were found" part | 18:37 |
| replaceafill | probably reset BIOS date | 18:37 |
| th1a | Hm. | 18:37 |
| replaceafill | that's why you see the 2014 in the db date | 18:37 |
| replaceafill | i want to go through the transaction history a bit | 18:37 |
| replaceafill | maybe that's what breaks things | 18:38 |
| replaceafill | in 32 | 18:38 |
| replaceafill | also | 18:38 |
| replaceafill | i'm not sure if Steph is confused about the testing instance | 18:38 |
| replaceafill | "I just ignored this and then carried on and was able to open schooltool using the link you sent through" | 18:39 |
| replaceafill | not sure what she means | 18:39 |
| replaceafill | can you follow that up? | 18:39 |
| replaceafill | or should we wait until i try transactions? | 18:39 |
| th1a | I'll follow up. | 18:40 |
| replaceafill | cool, thanks | 18:40 |
| th1a | I don't know if we have the exact cron setup they're all using. | 18:41 |
| th1a | I have trouble putting my hands on Fran's documentation... oh, is it in Dropbox? | 18:41 |
| replaceafill | yeah | 18:41 |
| th1a | OK. | 18:42 |
| th1a | At the risk of having them touch that whole area, I'm not sure they shouldn't have backups every six hours or so | 18:42 |
| replaceafill | it would make sense for rising | 18:43 |
| replaceafill | or set ups where they turn severs off | 18:43 |
| th1a | It would make sense for all of them -- there's no real downside. | 18:43 |
| replaceafill | right | 18:43 |
| th1a | 0 22 * * 5 repozo -R -v -r /home/administrator/backups -o /var/lib/schooltool/Data.fs | 18:45 |
| th1a | This is 10:00 weeknights? | 18:45 |
| th1a | 10:00 PM | 18:45 |
| replaceafill | yes | 18:46 |
| th1a | So it could just be 0 * * * * | 18:47 |
| th1a | At the top of every hour. | 18:47 |
| replaceafill | that'd work | 18:48 |
| th1a | And the whole point of repozo is that the web server is still up, right. | 18:48 |
| replaceafill | it's not like repozo takes a lot of time to run | 18:48 |
| th1a | Yeah. | 18:48 |
| th1a | Incrementally. | 18:49 |
| th1a | OK. I'll write a post on that subject. | 18:50 |
| replaceafill | cool | 18:51 |
| replaceafill | i'm starting the transaction investigation | 18:51 |
| replaceafill | i need to document this process... | 18:52 |
| th1a | I wonder if somehow just packing this on your system might help. | 18:52 |
| replaceafill | hhmm | 18:53 |
| replaceafill | i could try tht | 18:53 |
| replaceafill | that | 18:53 |
| th1a | It is easy! | 18:53 |
| replaceafill | i could reproduce the issue in my 32 bit VM | 18:53 |
| replaceafill | i mean, i was able | 18:53 |
| replaceafill | i'll try | 18:53 |
| th1a | Yes. | 18:54 |
| th1a | OK. | 18:54 |
| th1a | I guess that's it then. | 18:54 |
| replaceafill | ok | 18:55 |
| th1a | Thanks replaceafill. | 18:55 |
| replaceafill | thanks th1a | 18:55 |
| *** robb_nl has quit IRC | 19:01 | |
| th1a | replaceafill: aty? | 19:07 |
| replaceafill | yes | 19:08 |
| th1a | Should we have them run 'uname -a' ? | 19:08 |
| replaceafill | yeah, probably | 19:08 |
| replaceafill | the more we know about that server the better | 19:08 |
| replaceafill | also 'date' | 19:08 |
| th1a | OK. | 19:08 |
| th1a | replaceafill: Did we ever discuss the bug reports about 64-bit issues in the BTree implementation? | 19:11 |
| replaceafill | no | 19:12 |
| replaceafill | there are bug reports about that? | 19:12 |
| replaceafill | don't remember seeing them | 19:12 |
| th1a | https://mail.zope.org/pipermail/zodb-dev/2004-November/008200.html | 19:12 |
| th1a | Maybe I chose the right search terms this time. | 19:12 |
| replaceafill | oh boy | 19:14 |
| replaceafill | i'll avoid that for now... | 19:15 |
| th1a | does that look possible? | 19:15 |
| replaceafill | do you think that could be the cause for this? | 19:15 |
| th1a | ? | 19:15 |
| th1a | I don't know. | 19:16 |
| replaceafill | let's first wait from them running 'uname -a' | 19:16 |
| th1a | OK. | 19:17 |
| th1a | replaceafill: Maybe also you should just try moving a standard, working db from a 64 bit to 32 bit system. | 19:24 |
| replaceafill | ok | 19:26 |
| replaceafill | th1a, confirmed | 19:29 |
| replaceafill | a "working" db shows the same problem in 32 | 19:29 |
| th1a | OK, we can never move that way? | 19:29 |
| th1a | OK. | 19:29 |
| replaceafill | so, i should stop chasing transactions | 19:30 |
| th1a | I guess we could also be using anacron with cron to deal with the computer being off. | 19:31 |
| replaceafill | does that work when you turn it on? | 19:32 |
| th1a | Yes. | 19:32 |
| replaceafill | th1a, for the record moving from 32 to 64 doesn't work either | 19:46 |
| th1a | ok | 19:46 |
| th1a | We should use the full --recover flag instead of -R in the cron jobs too... | 20:03 |
| th1a | I guess I can add more people to the Launchpad collaboration team as long as one person is the Approver. | 20:13 |
| replaceafill | th1a, i think it's just the date | 21:41 |
| replaceafill | i'm going to get a quick lunch and try it | 21:41 |
| replaceafill | th1a, yep that's it | 22:05 |
| th1a | Huh... | 22:14 |
| th1a | Why did you have the 32/64 issue then? | 22:14 |
| th1a | replaceafill? | 22:14 |
| th1a | Oh, that is a separate issue. | 22:15 |
| replaceafill | th1a, who knows | 22:15 |
| replaceafill | yeah | 22:15 |
| th1a | OK. | 22:15 |
| th1a | I'll have to add a warning in the book. | 22:15 |
| replaceafill | thank god i have the disk error in my laptop every time i disconnect it | 22:16 |
| replaceafill | that's what really hinted me | 22:16 |
| th1a | Ah. | 22:16 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!