Evergreen ILS Website

IRC log for #evergreen, 2022-08-26

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
07:01 collum joined #evergreen
07:05 rjackson_isl_hom joined #evergreen
07:36 mantis1 joined #evergreen
07:46 BDorsey joined #evergreen
08:07 rfrasur joined #evergreen
08:32 rjackson_isl_hom joined #evergreen
08:38 mmorgan joined #evergreen
09:39 Dyrcona joined #evergreen
10:41 tlittle joined #evergreen
11:06 jihpringle joined #evergreen
11:14 mixo joined #evergreen
11:14 mixo Hello
11:15 mixo I am trying to install evergreen on debian 11
11:16 mixo I get an error: Package python-nose is not available, but is referred to by another package.
11:16 mixo This may mean that the package is missing, has been obsoleted, or
11:16 mixo is only available from another source
11:16 mixo E: Unable to locate package python-coverage
11:16 mixo E: Unable to locate package python-libxml2
11:16 mixo E: Unable to locate package python-memcache
11:17 mixo what can I do?
11:29 Dyrcona Install a more recent version of Evergreen. It looks like the one you're trying to install is missing the commit to remove the Python dependencies.
11:48 mixo Sorry more precisely I get this error when I try to install opensrf-3.2.2
11:49 mixo There is no later version
11:50 Dyrcona mixo: Oh, OK. Install OpenSRF from a git clone. I don't think that has made it to a tarball, yet.
11:50 mixo OK. Thank you.
11:51 Dyrcona Yeah, the fix is in the rel_3_2 branch in git.
11:51 Bmagic mixo: what Dyrcona said. Those python packages that you mention are still referenced in the OpenSRF install code, but are no longer available in the debian repo's
11:51 Dyrcona Or, you can install the master branch. It works with all current Evergreen releases.
11:52 Bmagic we use this trick sed -i '/python-memcache/ s/python-memcache//' src/extras/Makefile.install && sed -i '/python-pyxmpp/ s/python-pyxmpp//' src/extras/Makefile.install
11:53 * Dyrcona uses Git.
11:56 mixo Thank you very much
12:22 Dyrcona @decide cheesecake or peach pie
12:22 pinesol Dyrcona: go with peach pie
12:22 Dyrcona pinesol: I believe I will.
12:22 pinesol Dyrcona: Your computer account is overdrawn. Please reauthorize.
12:22 * mmorgan will take the cheesecake!
12:26 * Dyrcona totally forgot about the Bertucci's pizza dough rolls!
12:26 Dyrcona @dessert mmorgan
12:26 * pinesol grabs some Browser Cookies for mmorgan
12:26 Dyrcona heh
12:26 mmorgan :)
12:26 mmorgan Thanks for the thought, anyway!
12:27 mmorgan Those pizza dough rolls are the best!
12:28 Dyrcona Yeah we went to Bertucci's last night and they gave us some to take home with out leftovers.
12:29 Dyrcona mmorgan: It's caramel turtle cheesecake from last weekend. It's probably still good. :)
12:29 Dyrcona The peach pie is from Mann's Orchard in Methuen, and it's only a couple of days old.
12:29 mmorgan Dyrcona: Have some of each!
12:30 Dyrcona :)
12:30 Dyrcona Maybe I'll have some cheesecake later.
13:02 collum joined #evergreen
13:10 jihpringle joined #evergreen
13:15 Dyrcona So, I'm going to recommend against a self-hosted GitLab for the community. I've had some minor problems with it before, but after updating yesterday, I can't access repositories, gitlab-ctl status says everything is OK. I've searched the various error messages and nothing suggested works so far.
13:16 Dyrcona I'm going to see if I can dump/backup the repos. I doubt it because GitAly seems to be the busted component.--Good thing we (CWMARS) are moving to GitHub.
13:17 Dyrcona gitlab-ctl fixit doesn't exist. :)
13:19 Dyrcona Looks like a GitLab backup would be useless without another GitLab instance anyway.
13:20 Dyrcona And then there is this: Backup::Error: gitaly-backup exit status 1
13:33 Dyrcona There's too much going on for a Friday....
13:42 Dyrcona typos--
13:42 Dyrcona @karma typos
13:42 pinesol Dyrcona: Karma for "typos" has been increased 0 times and decreased 2 times for a total karma of -2.
13:43 Dyrcona The typos-- is not related to the GitLab thing, by the way. IDK what's wrong there....
13:50 Dyrcona Hmm. Be nice if git had a remote command to disable skip or ignore a remote so that I could leave it configured but just ignore it for a while.
14:00 Dyrcona mmorgan++ # For linking an "old" email chain to a bug. I had forgotten that discussion.
14:16 mmorgan @karma tyops
14:16 pinesol mmorgan: tyops has neutral karma.
14:16 mmorgan worth a shot.
14:21 Dyrcona :)
14:34 Dyrcona OK. Whatever you say, GitHb: "Great repository names are short and memorable. Need inspiration? How about upgraded-succotash?"
14:35 Dyrcona Here, I was gonna go with vmsetup.
14:35 berick sufferin-succotash
14:39 Dyrcona berick++
15:15 * Dyrcona hears thunder and smells rain.
15:16 mmorgan Rain? What's that?
15:18 Dyrcona Yeah, I know....
15:18 * Dyrcona waits for the dog to wake up and go nuts over the ear-shplitten-louden-boomers.
15:18 * mmorgan hears the thunder, too, now.
15:19 Dyrcona Ah, well, back to updating local git repos to point to GitHub.
15:19 Dyrcona too_many_servers--
15:19 mmorgan The pond behind our office building is almost dry. I've never seen it so low.
15:21 Dyrcona Yeah....
15:21 Dyrcona Pointed one repo at the wrong remote.....
15:37 Christineb joined #evergreen
15:54 Dyrcona So, even if you lose your main git repository, all of the local clones are basically backups. You can "recover" some lost branches from the local clones if the branches are still hanging around.
16:00 berick git clone --mirror and git remote update are also pretty handy for full backups
16:02 Dyrcona berick++ I'll look into those commands later.
16:08 Dyrcona cssh++
17:08 mmorgan left #evergreen
17:19 Bmagic berick: since the autorenewals essentially are two triggers, one for the renewal and one for the notice. How do you handle that with messagebee? I let your code do the whole thing, but it's looking like when it runs, it only runs the renewal but not the notice
17:20 Bmagic the notice trigger runs the next day, and for items that could not be renewed, the message about it not getting renewed is a bit stale
17:21 Bmagic I'm thinking about setting up a cron to run a few hours before your code does, to take care of the renewal and trigger the message be reactor to queue up. Then your code would run those maybe?
17:22 berick we autorenew at 2am and generate the notices at 8am currently.
17:22 berick generate-notices.sh --send-xml --granularity Auto-Renew-Email --end-date $(date --date '+1 day' +'\%F') --file-date $(date +'\%F')
17:22 berick escaped for cron
17:22 Bmagic ok, so, are both handled by ./generate-notices.sh ?
17:23 berick no, autorenew is just a regular A/T
17:23 Bmagic there we go, thanks! That's what I was thinking
17:24 Bmagic so, just to be safe, I might put a delay on those secondary triggers, like an hour? So that the regular A/T doesn't end up running them before generate-notices.sh gets a chance
17:28 berick Bmagic: probably better to make sure the notice generator only ever runs after all auto renew processing is complete
17:28 berick assuming you notify daily, a delay would just push notices back a day for any that took too long.
17:30 Bmagic right, makes sense. the timing of the cron is the main thing, if I set delay='3 hours' for the trigger with hook='autorenew' then the standard A/T will insert a row with a future run_time of 3 hours from now. And as long as the generate_notices.sh is run more than 3 hours from then, we're good
17:30 berick ah i give all my notices of this type a '5 second' delay so that can't happen
17:30 berick and just set the script to run when I know it's OK to send all the notices
17:31 Bmagic and with 5 seconds delay, the original A/T trigger doesn't run it?
17:31 Bmagic it just makes the entry in action_trigger.event but leaves it for another cron to pick it up?
17:31 berick not if you give the notice A/T a granularity
17:31 berick the generic A/T runner skips events that have a granularity
17:32 berick s/events/event defs/
17:32 Bmagic they have granularity
17:32 berick then they won't fire until you run generate-notices.sh --granularity XYZ ...
17:32 Bmagic action_trigger_runner.pl --process-hooks --run-pending --granularity messagebee-autorenew --granularity-only
17:33 Bmagic for example ^^^^ would both run the renewal AND any hook='autorenewal' at the same time?
17:34 Bmagic I should take run-pending out of that?
17:34 berick don't give them the same granularity
17:34 berick e.g. i have Auto-Renew-Processor and Auto-Renew-Email granularities
17:34 Bmagic I learned a couple of years ago that you can't connect two triggers together with two different granularities
17:35 Bmagic at least that's what I think* I remember.
17:36 berick hm, not ringing any bells
17:36 berick maybe if you wanted to run them at the same time?  not sure
17:36 Bmagic When the first one fires, it looks for other definitions within the same granularity. So the original trigger def is hook='Circ::AutoRenew' and the secondary is hook='autorenewal'. Both of those definitions don't have to be the same granularity?
17:38 berick yes, different granularities.  I think what makes it OK is the 2nd event-def is not running in real time, we're just creating events for it
17:38 berick that will run later w/ generate-notices.sh
17:38 Bmagic I'll give it a whirl!
17:38 Bmagic berick++

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat