Evergreen ILS Website

IRC log for #evergreen, 2018-01-11

| 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
00:18 Jillianne joined #evergreen
06:31 pinesol_green News from qatests: Test Failure <http://testing.evergreen-ils.org/~live>
07:29 rjackson_isl joined #evergreen
07:32 sandbergja joined #evergreen
07:32 agoben joined #evergreen
07:47 rlefaive joined #evergreen
08:29 jvwoolf joined #evergreen
08:45 mmorgan joined #evergreen
08:49 Dyrcona joined #evergreen
08:49 Dyrcona Whee! https://www.bleepingcomputer.com/news/sof​tware/meltdown-and-spectre-patches-causin​g-boot-issues-for-ubuntu-16-04-computers/
08:50 Dyrcona Though it looks like they may have fixed it already.
08:50 Dyrcona Too much confusion of the two issues. You can't really patch Spectre.
08:56 rhamby Dyrcona: time to break out the Amigas!
08:57 Dyrcona rhamby++
08:57 Dyrcona MC60x0 was a nice processor series.
08:58 Dyrcona Oops, missed the 8... 680x0
08:59 csharp Dyrcona: yeah, they released a working kernel (4.4.0-109) yesterday and I can confirm that it works
08:59 csharp embarassing for Ubuntu, tho
09:03 csharp I asked this the other day during off hours and didn't see (possibly missed) an answer: should we be using Hatch for more than just printing?
09:05 Dyrcona I don't think so. I think it's only other purpose was going to be offline circulation, and that isn't the case any more.
09:06 Dyrcona berick or JBoyer would probably know for sure.
09:12 csharp we're currently under the impression that there are advantages for persistent storage when using Hatch that aren't there otherwise
09:13 csharp but I'm remembering a discussion between JBoyer and Bmagic about it that sounded like that's a bad thing
09:20 Bmagic good morning
09:20 Bmagic Scrolling back, it looks like I have been chosen to bring some printers to the conference?
09:21 csharp Bmagic: only if it's convenient! :-)
09:22 Bmagic csharp: we are turning on three checkmarks for Hatch
09:22 csharp Bmagic: and all is well?
09:22 csharp (we're following your docs for the most part, btw)
09:23 Bmagic It puts the settings in Hatch instead of the browser and in theory is less volatile
09:23 Bmagic What is the problem with that?
09:23 csharp Bmagic: ok - great - that's in line with what we're telling the libs
09:24 Bmagic csharp: which docs?
09:24 csharp there has been a lot of confusion about what Hatch is for vs. what it was originally envisioned to be
09:24 csharp Bmagic: terran has some docs from MOBIUS for managing hatch setup
09:25 Bmagic probably http://libraries.missourievergreen.org​/content/staff-client-troubleshooting
09:26 csharp yes, that's it
09:26 csharp I've only seen them printed out on Terran's desk
09:28 yboston joined #evergreen
09:28 JBoyer I don't have any special knowledge about Hatch plans, but I had thought that when Lovefield went in for offline circ that it came up that Hatch may eventually be printing only for simplicity's sake.
09:29 csharp JBoyer: thanks
09:31 JBoyer (Though I do very much like the idea of "this is a small extension app to simplify printing" vs "this is how we do anything inconvenient to do in a web browser" which it isn't yet, but could become.)
09:32 * csharp agrees
09:32 csharp being from the pre-systemd UNIX world ;-)
09:41 Dyrcona csharp: Come to the BSD-side. We don't have systemd. :)
10:01 berick csharp: the consensus as I understand it is to move toward printing only for Hatch.  However, until we move workstation settings onto the server, Hatch can be useful for not accidentially blowing away localStorage.  well, it doesn't solve the problem, but it makes recovery easier.
10:02 berick fwiw, i'm strongly in favor of printing only for hatch
10:02 berick once the workstation stuff is done
10:02 berick serve-side ws settings and server-configured grid columns will be big priorities for me in the coming months.
10:12 rlefaive joined #evergreen
10:23 Dyrcona jeffdavis: I tried your odapi-checker.pl this morning.
10:23 Dyrcona jeffdavis: I get this error: Error on API request: 405 Method Not Allowed
10:23 Dyrcona That's on the test server.
10:23 Dyrcona Any idea what's wrong?
10:24 pinesol_green [evergreen|Jason Boyer] LP1741072: Fix JS test for template conversion - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ce01eeb>
10:24 pinesol_green [evergreen|Bill Erickson] LP#1741072 Volcopy editor deposit amount format repairs - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=2f7449a>
10:25 JBoyer Yay! I was wondering when those would go in.
10:26 JBoyer gmcharlt++
10:26 Dyrcona jeffdavis: If I try with production values, I get 400 bad request.
10:29 Bmagic I had a library come to me with a search result issue and I am wondering if bug 1736419 fixes it. The complaint is searching with preferred library X results with details turned on, not showing X's holdings in the copy table below each bib result.
10:29 pinesol_green Launchpad bug 1736419 in Evergreen "Search Showing Bibs with no Holdings" [High,Fix committed] https://launchpad.net/bugs/1736419
10:30 berick gmcharlt_: before I open a bug, do you know if we have anything written down re: storing workstation settings on the server?  nothing in the hackway gdoc, just a bullet on the agenda.
10:31 gmcharlt_ berick: no, I need to write that up
10:32 pinesol_green [evergreen|Cesar Velez] LP#1739786 - Webstaff MarcEditor BibSource Drop Down Order - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=36d72cf>
10:32 berick gmcharlt_: *nod* let me know if/how I can help.  i'd like to dig into that soon
10:33 gmcharlt joined #evergreen
10:38 Dyrcona jeffdavis: Just the basic check of the account works on the standard endpoint. I get 200 OK.
10:39 Dyrcona jeffdavis: Trying the same on the integration endpoint with our test account id gives 403 Forbidden, so likely something wrong on their end or I need a different token for testing that no one told me about?
10:40 pinesol_green [evergreen|Jason Boyer] LP1737052: Fix Typo in Permission Name - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=f3bff3e>
10:44 pinesol_green [evergreen|Jane Sandberg] LP1719943: fixing typo in credential testing interface - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=ce4b51e>
11:11 Bmagic Before I submit this on LP, has it already been reported that the acqusitions interface is bugged when adding a note/alert on an item. The dropdown UI shows up and immediately disappears?
11:14 csharp berick: thanks - that jibes with everything we're saying
11:14 csharp hooray!
11:29 miker JBoyer: re lovefield, that's there because hatch was not (still is not) ready to handle table-ish data, and it the support infrastructure for hatch was in heavy flux while I was working on offline, worrying me about availability and features for 3.0.0
11:29 miker (for the room, really, not just JBoyer)
11:30 Dyrcona Beautiful...More dropped log messages to the syslog server. I registered two workstations this morning and only see 1.
11:31 JBoyer miker++ # for the EXPLAIN ANALYZE, heh.
11:33 miker :)
11:36 Dyrcona rsyslog is next to useless.
11:46 rlefaive joined #evergreen
11:49 kmlussier joined #evergreen
11:52 kmlussier Has anyone noticed a problem in the web client where titles in the holds pull list disappear when you sort the list?
11:53 kmlussier I hadn't noticed it on my test VMs, but I saw it today when looking at a system with production data on it. https://drive.google.com/file/d/1F8​wTMwk9tcHKMvOUPj4UkFLs-67uFeiV/view
11:53 jeff Dyrcona: rsyslog problems?
11:53 kmlussier Oh, I guess that video isn't ready for prime time yet.
11:53 Dyrcona jeff: Things are not making it to the server, and it's seemingly random.
11:54 Dyrcona jeff: I made some tweaks suggested by JBoyer, but still some messages go missing.
11:54 Dyrcona If it is something recent, it's better to grep the brick head logs simultaneously.
11:57 jeff Dyrcona: if you haven't already, you might use the rsyslog daemon on the app servers then the omrelp output module to pass the messages to the actual syslog server/servers.
11:59 jeff (rsyslog on both ends, if that was unclear)
12:04 Dyrcona I believe it already is rsyslog on both ends, but too much to do and too little time.
12:06 jeff *nod*
12:07 khuckins joined #evergreen
12:08 dwgreen joined #evergreen
12:09 khuckins_ joined #evergreen
12:11 phasefx btw, it was a perl live test that failed this morning, not the browser build
12:12 phasefx Failed test 'Hold 263 has 31 mapped potential copies' at live_t/20-hold-targeter.t line 100. got: '24' expected: '31'
12:20 terran joined #evergreen
12:34 rlefaive joined #evergreen
12:49 jeffdavis Dyrcona: Overdrive client auth uses the same endpoint for both integration and production. IIRC you use the same token for both.
12:50 Dyrcona jeffdavis: So, my problem is believing their documentation.... :)
12:55 jeffdavis What were you doing that gave the 400 Bad request error?
12:57 Dyrcona Patron auth with our production account on the man server.
12:57 Dyrcona s/man/main/
12:59 Dyrcona The testing account worked for a basic connection without specifying the testing endpoint.
13:02 Dyrcona I get a 400 bad request trying patronauth with the testing credentials, too. I guess they didn't actually enable that for us.
13:09 jeffdavis I'm getting 200 OK for patron auth with production data. I don't have testing access set up at the moment so can't test that.
13:10 kipd Tuning question about the output from osrf_control --diagnostic:  I assume this isn't good, but is changing max_children for some of these processes the way to go?
13:10 Dyrcona Well, I'm supposed to have it for testing but not production, at least that's what I understood from the emails.
13:10 kipd https://www.irccloud.com/pastebin/Xo4UJYH6/
13:11 Dyrcona kipd: Yes, if you need more. It depends, really.
13:11 kipd Everything beyond max_children is waiting in queue right?
13:14 Dyrcona No. Everything beyond max_children is dropped on the floor.
13:14 kipd Ohhh, not good at all.
13:15 Dyrcona You'll get messages in the log when you run out. I was just looking for the wording of the message, but couldn't find it.
13:15 jeff kipd: are you trying to understand how to interpret the #drones field?
13:15 kipd Yes Jeff.
13:17 Dyrcona kipd: It's number of drones/max. configured
13:19 jeff i'm intrigued by open-ils.fielder's #drones=75/15 500%
13:19 jeff it does appear that the 15 is max_children for that service from the opensrf.settings service.
13:19 kipd Right, so the ones over the max configured are queued or dropped?
13:20 jeff well, the first number is intended to be the number of running drones -- which I think DOES include idle or spare drones.
13:20 kipd We’ve got pretty substantial hardware behind all of these so I’m thinking I have room to up those configured maximums but want to understand what the output really means.
13:21 jeff log entries indicating that you are hitting max_children would probably be a better indicator than "i am running the max number"
13:22 JBoyer One way to get numbers like that might be zombie processes, though that would be rather a lot of them.
13:22 Dyrcona yes, the first number includes the spares.
13:22 jeff but in theory you should never see more than 100% for #drones and I'm not sure what conditions would lead to that display (orphaned drones, something else...)
13:25 Dyrcona Yeah, I've never seen that output.
13:37 Dyrcona jeffdavis: Should I be making the test patron-auth requests against the standard endpoint or should I specify a test endpoint?
13:39 jeffdavis The standard one
13:39 jeffdavis "All patron authentication POSTs use the same URL (https://oauth-patron.overdrive.com/patrontoken) in both the production and integration environments" (http://developer.overdrive.com/apis/patron-auth)
13:40 Dyrcona Ah,ok.
13:40 Dyrcona I hadn't gotten that far into the docs.
13:41 Dyrcona I still get 400 bad request with testing parameters.
13:48 jeffdavis The --verbose option should give you the full response, which might help narrow down the cause. If nothing else you can pass the info along to their support people.
13:54 Dyrcona OK. I'll try that.
13:54 Dyrcona I'm starting to think that they just failed to enable patron auth.
13:55 Dyrcona Ah, looks like a problem on their end: There was an error determining which library authentication server to contact for your request.
14:05 Dyrcona Oh! I think I found something in an email from Overdrive that I overlooked.
14:06 Dyrcona Details! But I think I have it fixed.
14:06 Dyrcona I was using the wrong authorization name.
14:09 Dyrcona So, they sent me two library names in the email, the one that I'm supposed to use for the authorization name and the one that we choose on the test site. I missed the first more or less, 'cause it was on the same line with the other.
14:14 Dyrcona jeffdavis++ # Thanks for the help. I think I have it working, more or less.
14:15 jeffdavis yay!
14:17 pinesol_green [evergreen|Dan Wells] LP#1730470 Restore XUL serial receive compatibility - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=d90f69f>
14:19 Dyrcona I'm getting null of null available on the only record I've checked, but at least that's better than nothing at all. :)
14:21 csharp there's a pun in there somewhere dying to get out
14:22 Dyrcona Yeah. :)
14:22 Dyrcona null is better than nothing. :)
14:22 Dyrcona Irony much? :)
14:32 Dyrcona jeffdavis: Do you know where'd I'd find the authorizationname on the developer site?
14:32 Dyrcona I'm trying to check if I have the correct one for production.
14:46 jeffdavis You have to ask them for it, it's not on their site anywhere afaik.
14:48 jeffdavis Dyrcona: ^
14:49 Dyrcona jeffdavis: That's what I was beginning to think. I have an email from last April with most of that information.
14:54 Dyrcona I see something that I checked out on the integration site in my E-Items Checked out list, so that's working.
14:56 Dyrcona The Download button doesn't work and there's nothing in the drop down next to ti.
14:59 sandbergja joined #evergreen
15:17 mmorgan1 joined #evergreen
15:32 Dyrcona I think the null of null happens because the records don't exist on the api testing server.
15:33 Dyrcona I found one that looks like it would be there that is in our catalog, but the ISBN numbers don't match.
15:33 Dyrcona We also have a record for an e-book and e-audio from Overdrive, but the integration site only has an e-book with a different ISBN from our e-book.
15:37 kmlussier Dyrcona: bug 1677813
15:37 pinesol_green Launchpad bug 1677813 in Evergreen "Ebook API: OverDrive download link non-functional in My Account" [Undecided,New] https://launchpad.net/bugs/1677813
15:37 Dyrcona kmlussier++ I forgot about that.
16:27 khuckins_ joined #evergreen
16:29 khuckins__ joined #evergreen
17:02 mmorgan joined #evergreen
17:02 mmorgan left #evergreen
17:04 khuckins_ joined #evergreen
17:17 jvwoolf joined #evergreen
17:32 jvwoolf joined #evergreen
17:51 book` joined #evergreen
18:31 pinesol_green News from qatests: Test Failure <http://testing.evergreen-ils.org/~live>
18:50 Jillianne joined #evergreen
23:24 jonadab joined #evergreen

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