<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en">
<head>
<meta name="description" content="IRC LOG for channel #evergreen" />
<meta name="viewport" content="width=device-width" />
<link rel="stylesheet" type="text/css" href="/s/style.css" title="Irclog Stylesheet" />
<link rel="shortcut icon" href="/s/favicon.ico" type="image/x-icon" />
<link rel="canonical"
href="/evergreen/2022-12-09" />
<title>IRC log for #evergreen, 2022-12-09</title>
</head>
<body>
<p><a href="http://evergreen-ils.org/"><img style="float:right; border:none"
src="/s/small_logo.jpg" alt="Evergreen ILS Website"/></a></p>
<h1>IRC log for #evergreen, 2022-12-09</h1>
<p>
<a href="/evergreen/2022-12-08" rel="prev">
← Previous day</a> |
<a href="/">Channels</a> |
<a href="/evergreen/">#evergreen index</a> |
<a href="/evergreen/today">Today</a> |
<a href="/evergreen/2022-12-10" rel="next">
Next day →</a> |
<a href="/evergreen/search">Search</a> |
<a href="http://www.google.com/search?q=site%3Airc.evergreen-ils.org+inurl%3Aevergreen">Google Search</a>
| <a href="/evergreen/2022-12-09/text">Plain-Text</a> |
<a href="/evergreen/2022-12-09/summary">summary</a>
| <a href="https://webchat.freenode.net/?channels=evergreen">Join Webchat</a>
</p>
<p>All times shown according to the server's local time.</p>
<div>
<span style="display: none;" id="filter_toggle"></span>
<span id="summary_container" />
</div>
<form action="#" name="summary_form" id="summary_form">
<span id="top" />
<table id="log" style="clear:both">
<tr class="head">
<th>Time</th>
<th>Nick</th>
<th>Message</th>
</tr>
<tr id="id_l2" class="new special dark">
<td class="time" id="i_517321"><a href="/evergreen/2022-12-09#i_517321">07:07</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">kworstell-isl joined #evergreen</td>
</tr>
<tr id="id_l3" class="cont special dark">
<td class="time" id="i_517322"><a href="/evergreen/2022-12-09#i_517322">08:23</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">JBoyer joined #evergreen</td>
</tr>
<tr id="id_l4" class="cont special dark">
<td class="time" id="i_517323"><a href="/evergreen/2022-12-09#i_517323">08:31</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">mantis1 joined #evergreen</td>
</tr>
<tr id="id_l5" class="cont special dark">
<td class="time" id="i_517324"><a href="/evergreen/2022-12-09#i_517324">08:42</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">mmorgan joined #evergreen</td>
</tr>
<tr id="id_l6" class="cont special dark">
<td class="time" id="i_517325"><a href="/evergreen/2022-12-09#i_517325">08:52</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">Stompro joined #evergreen</td>
</tr>
<tr id="id_l7" class="new nick nick_Bmagic">
<td class="time" id="i_517326"><a href="/evergreen/2022-12-09#i_517326">09:06</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">I can't seem to figure out where this index comes from: CREATE INDEX mrca_vlist_idx ON metabib.record_attr_vector_list USING gin (vlist evergreen.gin__int_ops);</td>
</tr>
<tr id="id_l8" class="cont nick nick_Bmagic">
<td class="time" id="i_517327"><a href="/evergreen/2022-12-09#i_517327">09:07</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">pg_restore throws that as an error. evergreen.gin_int_ops and evergreen.gin__int_ops doesn't appear in the Evergreen source</td>
</tr>
<tr id="id_l9" class="new special dark">
<td class="time" id="i_517328"><a href="/evergreen/2022-12-09#i_517328">09:08</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">Dyrcona joined #evergreen</td>
</tr>
<tr id="id_l10" class="new nick nick_Bmagic">
<td class="time" id="i_517329"><a href="/evergreen/2022-12-09#i_517329">09:08</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">wb Dyrcona</td>
</tr>
<tr id="id_l11" class="new nick nick_csharp_ dark">
<td class="time" id="i_517330"><a href="/evergreen/2022-12-09#i_517330">09:21</a></td>
<td style="color: #4a1700" class="nick">csharp_</td>
<td class="msg ">Bmagic: I see it in three files: <a href="https://pastebin.com/4aAnhmWg" title="https://pastebin.com/4aAnhmWg">https://pastebin.com/4aAnhmWg</a></td>
</tr>
<tr id="id_l12" class="new nick nick_Bmagic">
<td class="time" id="i_517331"><a href="/evergreen/2022-12-09#i_517331">09:22</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">csharp_++ # git grep ftw</td>
</tr>
<tr id="id_l13" class="new nick nick_csharp_ dark">
<td class="time" id="i_517332"><a href="/evergreen/2022-12-09#i_517332">09:22</a></td>
<td style="color: #4a1700" class="nick">csharp_</td>
<td class="msg ">Bmagic: just learned that from Dyrcona</td>
</tr>
<tr id="id_l14" class="new nick nick_Bmagic">
<td class="time" id="i_517333"><a href="/evergreen/2022-12-09#i_517333">09:22</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">it must have been dropped</td>
</tr>
<tr id="id_l15" class="new nick nick_csharp_ dark">
<td class="time" id="i_517334"><a href="/evergreen/2022-12-09#i_517334">09:22</a></td>
<td style="color: #4a1700" class="nick">csharp_</td>
<td class="msg ">or maybe gmcharlt - not sure which</td>
</tr>
<tr id="id_l16" class="new nick nick_Bmagic">
<td class="time" id="i_517335"><a href="/evergreen/2022-12-09#i_517335">09:23</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">I know there was a big switch from one type of index to another a couple versions ago</td>
</tr>
<tr id="id_l17" class="cont nick nick_Bmagic">
<td class="time" id="i_517336"><a href="/evergreen/2022-12-09#i_517336">09:29</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">so, trying to track down more information. I used: git log -p -- Open-ILS/src/sql/Pg/030.schema.metabib.sql</td>
</tr>
<tr id="id_l18" class="cont nick nick_Bmagic">
<td class="time" id="i_517337"><a href="/evergreen/2022-12-09#i_517337">09:29</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">I see the line: CREATE INDEX mrca_vlist_idx ON metabib.record_attr_vector_list USING gin ( vlist gin__int_ops );</td>
</tr>
<tr id="id_l19" class="new special dark">
<td class="time" id="i_517338"><a href="/evergreen/2022-12-09#i_517338">09:30</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">collum joined #evergreen</td>
</tr>
<tr id="id_l20" class="new nick nick_Bmagic">
<td class="time" id="i_517339"><a href="/evergreen/2022-12-09#i_517339">09:30</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">but nowhere in that history do I see CREATE INDEX mrca_vlist_idx ON metabib.record_attr_vector_list USING gin ( vlist evergreen.gin__int_ops );</td>
</tr>
<tr id="id_l21" class="cont nick nick_Bmagic">
<td class="time" id="i_517340"><a href="/evergreen/2022-12-09#i_517340">09:30</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">the difference being evergreen.gin__int_ops vs. gin__int_ops</td>
</tr>
<tr id="id_l22" class="new nick nick_Dyrcona dark">
<td class="time" id="i_517341"><a href="/evergreen/2022-12-09#i_517341">09:32</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Bmagic: gin__int_ops should not be in the evergreen schema.</td>
</tr>
<tr id="id_l23" class="cont nick nick_Dyrcona dark">
<td class="time" id="i_517342"><a href="/evergreen/2022-12-09#i_517342">09:33</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">It's part of the intarray extension.</td>
</tr>
<tr id="id_l24" class="cont nick nick_Dyrcona dark">
<td class="time" id="i_517343"><a href="/evergreen/2022-12-09#i_517343">09:33</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">And, this may help with our search issues on pg 12+, but I still have to look into that more today.</td>
</tr>
<tr id="id_l25" class="cont nick nick_Dyrcona dark">
<td class="time" id="i_517344"><a href="/evergreen/2022-12-09#i_517344">09:35</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Bmagic: So, try to alter the index or drop it and recreate it with the line from the schema files. If I had to guess, I suspect something got botched during an upgrade.</td>
</tr>
<tr id="id_l26" class="new nick nick_Bmagic">
<td class="time" id="i_517345"><a href="/evergreen/2022-12-09#i_517345">09:39</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">After the restore finished, I manually added the index without the evergreen qualifier, and it worked. That's not a problem. I was just diving deeper into the git history to try and figure out when* we updated our code to drop that qualifier off the CREATE INDEX statement</td>
</tr>
<tr id="id_l27" class="cont nick nick_Bmagic">
<td class="time" id="i_517346"><a href="/evergreen/2022-12-09#i_517346">09:41</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">so far, I've not been able to track it down. I assume it would appear in 030.schema.metabib.sql as evergreen.gin__int_ops at some point in history</td>
</tr>
<tr id="id_l28" class="cont nick nick_Bmagic">
<td class="time" id="i_517347"><a href="/evergreen/2022-12-09#i_517347">09:42</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">maybe postgres stuck that on there without us telling it to</td>
</tr>
<tr id="id_l29" class="new nick nick_Dyrcona dark">
<td class="time" id="i_517348"><a href="/evergreen/2022-12-09#i_517348">09:43</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">That's because the evergreen qualifier was never there.</td>
</tr>
<tr id="id_l30" class="new nick nick_Bmagic">
<td class="time" id="i_517349"><a href="/evergreen/2022-12-09#i_517349">09:44</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">I'm slowly realizing that.</td>
</tr>
<tr id="id_l31" class="cont nick nick_Bmagic">
<td class="time" id="i_517350"><a href="/evergreen/2022-12-09#i_517350">09:44</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">So, why in the world would PG have it like that?</td>
</tr>
<tr id="id_l32" class="new nick nick_Dyrcona dark">
<td class="time" id="i_517351"><a href="/evergreen/2022-12-09#i_517351">09:45</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">grep -r intarray Open-ILS/src/sql/Pg might help provide a clue.</td>
</tr>
<tr id="id_l33" class="cont nick nick_Dyrcona dark">
<td class="time" id="i_517352"><a href="/evergreen/2022-12-09#i_517352">09:46</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">It could be that when one of the upgrade scripts was run to add intarray, the extension was added to the evergreen schema rather than the public schema.</td>
</tr>
<tr id="id_l34" class="new nick nick_Bmagic">
<td class="time" id="i_517353"><a href="/evergreen/2022-12-09#i_517353">09:46</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">I'll buy that</td>
</tr>
<tr id="id_l35" class="new special dark">
<td class="time" id="i_517354"><a href="/evergreen/2022-12-09#i_517354">09:47</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">dguarrac joined #evergreen</td>
</tr>
<tr id="id_l36" class="new nick nick_Dyrcona">
<td class="time" id="i_517355"><a href="/evergreen/2022-12-09#i_517355">09:55</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Talking to a read-only database: ERROR: canceling statement due to conflict with recovery</td>
</tr>
<tr id="id_l37" class="cont nick nick_Dyrcona">
<td class="time" id="i_517356"><a href="/evergreen/2022-12-09#i_517356">09:57</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">OK, fine, but then try again in a couple of minutes: SSL connection has been closed unexpectedly. The connection to the server was lost. Attempting reset: Succeeded.</td>
</tr>
<tr id="id_l38" class="cont nick nick_Dyrcona">
<td class="time" id="i_517357"><a href="/evergreen/2022-12-09#i_517357">09:57</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Guess that's not catastrophic after all.</td>
</tr>
<tr id="id_l39" class="cont nick nick_Dyrcona">
<td class="time" id="i_517358"><a href="/evergreen/2022-12-09#i_517358">10:03</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker: To answer your question from yesterday about how many bibs we have: deleted='t' 1,985,205; deleted='f' 2,237,018. This is the database used to generate the explain output on the different Pg releases. It's a dump of our production database from 12/04 and upgraded to 3.10.0.</td>
</tr>
<tr id="id_l40" class="cont nick nick_Dyrcona">
<td class="time" id="i_517359"><a href="/evergreen/2022-12-09#i_517359">10:18</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">It looks like it is the where on bre.vis_attr_vector. I just tried 'select id from biblio.record_entry where not ( int4range(0,268435455,'[]') @> ANY(vis_attr_vector);' and it is still running.</td>
</tr>
<tr id="id_l41" class="cont nick nick_Dyrcona">
<td class="time" id="i_517360"><a href="/evergreen/2022-12-09#i_517360">10:19</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I'm going to add in the NOT NULL and run it again with explain after it returns.</td>
</tr>
<tr id="id_l42" class="cont nick nick_Dyrcona">
<td class="time" id="i_517361"><a href="/evergreen/2022-12-09#i_517361">10:20</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Sorry, IS NULL rather than NOT NULL.</td>
</tr>
<tr id="id_l43" class="cont nick nick_Dyrcona">
<td class="time" id="i_517362"><a href="/evergreen/2022-12-09#i_517362">10:35</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Nope....</td>
</tr>
<tr id="id_l44" class="cont nick nick_Dyrcona">
<td class="time" id="i_517363"><a href="/evergreen/2022-12-09#i_517363">10:36</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Well, that took 3.7 seconds.</td>
</tr>
<tr id="id_l45" class="cont nick nick_Dyrcona">
<td class="time" id="i_517364"><a href="/evergreen/2022-12-09#i_517364">10:38</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Not bad though, 3.7 seconds for a seq scan on 3.7 million records.</td>
</tr>
<tr id="id_l46" class="new nick nick_Bmagic dark">
<td class="time" id="i_517365"><a href="/evergreen/2022-12-09#i_517365">10:39</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">better than a human</td>
</tr>
<tr id="id_l47" class="new nick nick_Dyrcona">
<td class="time" id="i_517366"><a href="/evergreen/2022-12-09#i_517366">10:43</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Adding an index on bre.vis_attr_vector doesn't make a difference in the explain output for the simple query.</td>
</tr>
<tr id="id_l48" class="cont nick nick_Dyrcona">
<td class="time" id="i_517367"><a href="/evergreen/2022-12-09#i_517367">10:44</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Also, there's an error in the SQL I pasted before. I'm actually running this to test: select id from biblio.record_entry where vis_attr_vector IS NULL OR NOT int4range(0,268435455,'[]') @> ANY(vis_attr_vector);</td>
</tr>
<tr id="id_l49" class="cont nick nick_Dyrcona">
<td class="time" id="i_517368"><a href="/evergreen/2022-12-09#i_517368">10:44</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I'm going to add an index on acvac.vis_attr_vector and try the big query again.</td>
</tr>
<tr id="id_l50" class="cont nick nick_Dyrcona">
<td class="time" id="i_517369"><a href="/evergreen/2022-12-09#i_517369">10:47</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">That doesn't seem to make a difference either, so I'll try the procedural index that miker suggested yesterday next.</td>
</tr>
<tr id="id_l51" class="cont nick nick_Dyrcona">
<td class="time" id="i_517370"><a href="/evergreen/2022-12-09#i_517370">10:54</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">"create index bre_not_deleted_idx ON biblio.record_entry (id) where deleted = 'f';" doesn't seem to help, either.</td>
</tr>
<tr id="id_l52" class="new nick nick_miker dark">
<td class="time" id="i_517371"><a href="/evergreen/2022-12-09#i_517371">10:59</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: since the EXPLAIN-reported predicate is (IIRC) "NOT deleted" you may want to try the index with that instead. but it may not help because an index is only used when it's expected to match less than a small part of the table</td>
</tr>
<tr id="id_l53" class="new nick nick_Dyrcona">
<td class="time" id="i_517372"><a href="/evergreen/2022-12-09#i_517372">11:04</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker: All right. I'll remake the index. I just dropped it and the other two to run the original again. I noticed a 300,000ms difference, but I want to rule out caching.</td>
</tr>
<tr id="id_l54" class="cont nick nick_Dyrcona">
<td class="time" id="i_517373"><a href="/evergreen/2022-12-09#i_517373">11:05</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Yeahp. Cache... The original with added indexes is finishing in the same time range.</td>
</tr>
<tr id="id_l55" class="cont nick nick_Dyrcona">
<td class="time" id="i_517374"><a href="/evergreen/2022-12-09#i_517374">11:06</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">create index bre_not_deleted_idx ON biblio.record_entry (id) where not deleted;</td>
</tr>
<tr id="id_l56" class="new nick nick_miker dark">
<td class="time" id="i_517375"><a href="/evergreen/2022-12-09#i_517375">11:09</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Bmagic: extensions can be created in specific schemas, and pre-extension contribs can be carried forward in upgrades. if you're restoring from a dump of a db where intarray (or the pre-extension contrib) was installed in the evergreen schema, but the db you're restoring into has it in public, you'll see what you're seeing</td>
</tr>
<tr id="id_l57" class="cont nick nick_miker dark">
<td class="time" id="i_517376"><a href="/evergreen/2022-12-09#i_517376">11:10</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">ah! I see Dyrcona came to the same conclusion/suggestion :)</td>
</tr>
<tr id="id_l58" class="new nick nick_Dyrcona">
<td class="time" id="i_517377"><a href="/evergreen/2022-12-09#i_517377">11:12</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker: Performance is roughly the same with the index (Execution Time: 114266.883 ms) as without (Execution Time: 114872.265 ms).</td>
</tr>
<tr id="id_l59" class="new nick nick_miker dark">
<td class="time" id="i_517378"><a href="/evergreen/2022-12-09#i_517378">11:14</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: re your issue, the other question I had was about the ingest.metarecord_mapping.preserve_on_delete internal flag. if that's true then the m.metarecord_source_map entries stick around on bib "delete", which is good if you want to be able to search for "deleted" bibs, but bad for this plan it would seem. is that on? (NTS: maybe we should check the flag and only use AND NOT deleted when it could have an effect?)</td>
</tr>
<tr id="id_l60" class="cont nick nick_miker dark">
<td class="time" id="i_517379"><a href="/evergreen/2022-12-09#i_517379">11:14</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">select * from config.internal_flag where name like '%delete%'; if you're in psql</td>
</tr>
<tr id="id_l61" class="new nick nick_Dyrcona">
<td class="time" id="i_517380"><a href="/evergreen/2022-12-09#i_517380">11:16</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I'm pretty sure that's on, but I'll double check.</td>
</tr>
<tr id="id_l62" class="cont nick nick_Dyrcona">
<td class="time" id="i_517381"><a href="/evergreen/2022-12-09#i_517381">11:17</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Nope. enabled = 'f'.</td>
</tr>
<tr id="id_l63" class="new nick nick_miker dark">
<td class="time" id="i_517382"><a href="/evergreen/2022-12-09#i_517382">11:17</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">kk</td>
</tr>
<tr id="id_l64" class="cont nick nick_miker dark">
<td class="time" id="i_517383"><a href="/evergreen/2022-12-09#i_517383">11:20</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: a big change in pg 12 is the default for non-materialized CTEs. we can't just do it across the board yet, but do you have the full original query handy to test by hand? if so, in front of each CTE (WITH-clause), after the AS keyword, what happens if you add MATERIALIZED between AS and the open-( ?</td>
</tr>
<tr id="id_l65" class="new nick nick_Dyrcona">
<td class="time" id="i_517384"><a href="/evergreen/2022-12-09#i_517384">11:21</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I'll give that a shot. I have it open in my editor.</td>
</tr>
<tr id="id_l66" class="new nick nick_miker dark">
<td class="time" id="i_517385"><a href="/evergreen/2022-12-09#i_517385">11:23</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">if that makes things happy -- and I can imagine it will because estimates of the number rows coming out of a tsearch CTE are ... usually way off -- then we /can/ add a PG version check and add it when it's available. there's precedent for that in the record attribute testing infrastructure.</td>
</tr>
<tr id="id_l67" class="new nick nick_Dyrcona">
<td class="time" id="i_517386"><a href="/evergreen/2022-12-09#i_517386">11:27</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">It's worse, but I may have missed a couple: Execution Time: 115367.485 ms</td>
</tr>
<tr id="id_l68" class="cont nick nick_Dyrcona">
<td class="time" id="i_517387"><a href="/evergreen/2022-12-09#i_517387">11:31</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">OK! I missed c_attr and b_attr before. after adding materialized to those two, it was much better: Execution Time: 3123.997 ms</td>
</tr>
<tr id="id_l69" class="cont nick nick_Dyrcona">
<td class="time" id="i_517388"><a href="/evergreen/2022-12-09#i_517388">11:32</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">My guess is c_attr as materialized does the magic.</td>
</tr>
<tr id="id_l70" class="cont nick nick_Dyrcona">
<td class="time" id="i_517389"><a href="/evergreen/2022-12-09#i_517389">11:33</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">For the sake of satisfying my curiosity, I'm going to try it with materialized just on c_attr.</td>
</tr>
<tr id="id_l71" class="new nick nick_miker dark">
<td class="time" id="i_517390"><a href="/evergreen/2022-12-09#i_517390">11:33</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">oh, interesting! but, not entirely surprising given what I /just/ saw in the docs at the end of this section: <a href="https://www.postgresql.org/docs/14/queries-with.html#id-1" title="https://www.postgresql.org/docs/14/queries-with.html#id-1">https://www.postgresql.org/docs/14/queries-with.html#id-1</a>.5.6.12.7 ("very_expensive_function()")</td>
</tr>
<tr id="id_l72" class="new nick nick_Dyrcona">
<td class="time" id="i_517391"><a href="/evergreen/2022-12-09#i_517391">11:34</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">"That's a bingo!"</td>
</tr>
<tr id="id_l73" class="new nick nick_miker dark">
<td class="time" id="i_517392"><a href="/evergreen/2022-12-09#i_517392">11:37</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">limit and offset are usually optimization fences ... Dyrcona, would you mind trying this? remove the MATERIALIZED on c_attr and add "OFFSET 0" just before the closing ) on that CTE</td>
</tr>
<tr id="id_l74" class="new nick nick_Dyrcona">
<td class="time" id="i_517393"><a href="/evergreen/2022-12-09#i_517393">11:38</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">OK. I was about to say that we'd have to bump the minimum Pg version to 12 if we use materialized CTEs.</td>
</tr>
<tr id="id_l75" class="new nick nick_miker dark">
<td class="time" id="i_517394"><a href="/evergreen/2022-12-09#i_517394">11:38</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">if that has the same effect, we should be able to (for now, until PG learns how to ignore useless limit/offset) do that.</td>
</tr>
<tr id="id_l76" class="cont nick nick_miker dark">
<td class="time" id="i_517395"><a href="/evergreen/2022-12-09#i_517395">11:39</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">as for versioning, we already test for the version in one other place and change query structure based on that. we could do it here if MATERIALIZED is the answer</td>
</tr>
<tr id="id_l77" class="cont nick nick_miker dark">
<td class="time" id="i_517396"><a href="/evergreen/2022-12-09#i_517396">11:40</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">we could also mark the function volatile to force materialization</td>
</tr>
<tr id="id_l78" class="new nick nick_Dyrcona">
<td class="time" id="i_517397"><a href="/evergreen/2022-12-09#i_517397">11:40</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Execution Time: 2675.806 ms</td>
</tr>
<tr id="id_l79" class="new nick nick_miker dark">
<td class="time" id="i_517398"><a href="/evergreen/2022-12-09#i_517398">11:40</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">oh hey, look at that</td>
</tr>
<tr id="id_l80" class="new nick nick_Dyrcona">
<td class="time" id="i_517399"><a href="/evergreen/2022-12-09#i_517399">11:40</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">So, the offset works, too.</td>
</tr>
<tr id="id_l81" class="new nick nick_miker dark">
<td class="time" id="i_517400"><a href="/evergreen/2022-12-09#i_517400">11:40</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">and, just to clarify, right now that offset is the /only/ change to the query?</td>
</tr>
<tr id="id_l82" class="new nick nick_Dyrcona">
<td class="time" id="i_517401"><a href="/evergreen/2022-12-09#i_517401">11:41</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Yes.</td>
</tr>
<tr id="id_l83" class="new nick nick_miker dark">
<td class="time" id="i_517402"><a href="/evergreen/2022-12-09#i_517402">11:41</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">right on</td>
</tr>
<tr id="id_l84" class="new nick nick_Dyrcona">
<td class="time" id="i_517403"><a href="/evergreen/2022-12-09#i_517403">11:42</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Which function were you thinking of marking volatile?</td>
</tr>
<tr id="id_l85" class="new nick nick_miker dark">
<td class="time" id="i_517404"><a href="/evergreen/2022-12-09#i_517404">11:43</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">search.calculate_visibility_attribute_test</td>
</tr>
<tr id="id_l86" class="new nick nick_Dyrcona">
<td class="time" id="i_517405"><a href="/evergreen/2022-12-09#i_517405">11:44</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">OK. I wasn't sure because asset.patron_default_visibility_mask is also used.</td>
</tr>
<tr id="id_l87" class="cont nick nick_Dyrcona">
<td class="time" id="i_517406"><a href="/evergreen/2022-12-09#i_517406">11:44</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Should I try that, too, or are we happy with "offset 0"?</td>
</tr>
<tr id="id_l88" class="new nick nick_miker dark">
<td class="time" id="i_517407"><a href="/evergreen/2022-12-09#i_517407">11:44</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">it is, but only in OPAC searches. staff search does use search.calculate_visibility_attribute_test thought</td>
</tr>
<tr id="id_l89" class="cont nick nick_miker dark">
<td class="time" id="i_517408"><a href="/evergreen/2022-12-09#i_517408">11:45</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">though, even</td>
</tr>
<tr id="id_l90" class="new nick nick_Dyrcona">
<td class="time" id="i_517409"><a href="/evergreen/2022-12-09#i_517409">11:45</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">FYI, I'm testing on Pg 13, but it probably doesn't matter.</td>
</tr>
<tr id="id_l91" class="new nick nick_miker dark">
<td class="time" id="i_517410"><a href="/evergreen/2022-12-09#i_517410">11:47</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">offset 0 is the more self-documenting variant that works for all pg version, for now. though, again, we can version-test and use materialized in 12+</td>
</tr>
<tr id="id_l92" class="cont nick nick_miker dark">
<td class="time" id="i_517411"><a href="/evergreen/2022-12-09#i_517411">11:49</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">testing offset 0 on an older version, just to be sure</td>
</tr>
<tr id="id_l93" class="new nick nick_Dyrcona">
<td class="time" id="i_517412"><a href="/evergreen/2022-12-09#i_517412">11:49</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I was talking about marking search.calculate_visibility_attribute_test as volatile.</td>
</tr>
<tr id="id_l94" class="new nick nick_miker dark">
<td class="time" id="i_517413"><a href="/evergreen/2022-12-09#i_517413">11:50</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">right</td>
</tr>
<tr id="id_l95" class="new nick nick_Dyrcona">
<td class="time" id="i_517414"><a href="/evergreen/2022-12-09#i_517414">11:50</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">I'm going to try it just because I can.</td>
</tr>
<tr id="id_l96" class="new nick nick_miker dark">
<td class="time" id="i_517415"><a href="/evergreen/2022-12-09#i_517415">11:50</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">I'm saying "offset 0" is more self-documenting, because the query has a marker of "something odd here" ... that function isn't /actually/ volatile</td>
</tr>
<tr id="id_l97" class="cont nick nick_miker dark">
<td class="time" id="i_517416"><a href="/evergreen/2022-12-09#i_517416">11:50</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">+1, try away!</td>
</tr>
<tr id="id_l98" class="new nick nick_Dyrcona">
<td class="time" id="i_517417"><a href="/evergreen/2022-12-09#i_517417">11:52</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Volatile without offset 0 works, too, but I'll defer to your opinion on offset 0.</td>
</tr>
<tr id="id_l99" class="cont nick nick_Dyrcona">
<td class="time" id="i_517418"><a href="/evergreen/2022-12-09#i_517418">11:53</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Execution time is about the same, 4ms difference which is probably not related to the change.</td>
</tr>
<tr id="id_l100" class="new nick nick_miker dark">
<td class="time" id="i_517419"><a href="/evergreen/2022-12-09#i_517419">12:17</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">I got curious and checked the difference between ts_rank and ts_rank_cd ... it's bigger than I thought :(</td>
</tr>
<tr id="id_l101" class="cont nick nick_miker dark">
<td class="time" id="i_517420"><a href="/evergreen/2022-12-09#i_517420">12:18</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">in my test query, 3ms vs 90ms, respectively</td>
</tr>
<tr id="id_l102" class="new special">
<td class="time" id="i_517421"><a href="/evergreen/2022-12-09#i_517421">12:22</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">jihpringle joined #evergreen</td>
</tr>
<tr id="id_l103" class="new nick nick_miker dark">
<td class="time" id="i_517422"><a href="/evergreen/2022-12-09#i_517422">12:28</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: do you want to branchify that, or shall I?</td>
</tr>
<tr id="id_l104" class="new nick nick_Dyrcona">
<td class="time" id="i_517423"><a href="/evergreen/2022-12-09#i_517423">12:28</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker: ts_rank makes a roughly 67ms difference in execution time.</td>
</tr>
<tr id="id_l105" class="cont nick nick_Dyrcona">
<td class="time" id="i_517424"><a href="/evergreen/2022-12-09#i_517424">12:28</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">in my test environment.</td>
</tr>
<tr id="id_l106" class="cont nick nick_Dyrcona">
<td class="time" id="i_517425"><a href="/evergreen/2022-12-09#i_517425">12:29</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">If you want to branch it, that's fine with me. I was going to open a Lp bug, but I should actually look into yet another OOM on the db server.</td>
</tr>
<tr id="id_l107" class="cont nick nick_Dyrcona">
<td class="time" id="i_517426"><a href="/evergreen/2022-12-09#i_517426">12:29</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">It happened just before noon.</td>
</tr>
<tr id="id_l108" class="new nick nick_miker dark">
<td class="time" id="i_517427"><a href="/evergreen/2022-12-09#i_517427">12:29</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">oh, well, that's not enough to warrant further investigation this moment. my test case may be unusual</td>
</tr>
<tr id="id_l109" class="cont nick nick_miker dark">
<td class="time" id="i_517428"><a href="/evergreen/2022-12-09#i_517428">12:29</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">I'll put together the patch and await your LP, then?</td>
</tr>
<tr id="id_l110" class="new nick nick_Dyrcona">
<td class="time" id="i_517429"><a href="/evergreen/2022-12-09#i_517429">12:30</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">OK with me.</td>
</tr>
<tr id="id_l111" class="cont nick nick_Dyrcona">
<td class="time" id="i_517430"><a href="/evergreen/2022-12-09#i_517430">12:36</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Is it possible to POST an OPAC search?</td>
</tr>
<tr id="id_l112" class="new nick nick_miker dark">
<td class="time" id="i_517431"><a href="/evergreen/2022-12-09#i_517431">12:42</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">it is</td>
</tr>
<tr id="id_l113" class="cont nick nick_miker dark">
<td class="time" id="i_517432"><a href="/evergreen/2022-12-09#i_517432">12:43</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">(and, note, i was incorrect about staff search not using the patron vis test function -- it does, it just ignores that the the WHERE clause)</td>
</tr>
<tr id="id_l114" class="cont nick nick_miker dark">
<td class="time" id="i_517433"><a href="/evergreen/2022-12-09#i_517433">12:43</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: <a href="https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/miker/optimization-fence-for-pg-12-CTEs-in-search" title="https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/miker/optimization-fence-for-pg-12-CTEs-in-search">https://git.evergreen-ils.org/?p=working/Evergreen.git;a=shortlog;h=refs/heads/user/miker/optimization-fence-for-pg-12-CTEs-in-search</a> for your eyeballs</td>
</tr>
<tr id="id_l115" class="new nick nick_Dyrcona">
<td class="time" id="i_517434"><a href="/evergreen/2022-12-09#i_517434">12:45</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker++</td>
</tr>
<tr id="id_l116" class="cont nick nick_Dyrcona">
<td class="time" id="i_517435"><a href="/evergreen/2022-12-09#i_517435">12:46</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Re POST: I guess we'll need to work out a mitigation for POSTS with "bad" queries in them.</td>
</tr>
<tr id="id_l117" class="new special dark">
<td class="time" id="i_517436"><a href="/evergreen/2022-12-09#i_517436">13:33</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">kworstell-isl joined #evergreen</td>
</tr>
<tr id="id_l118" class="cont special dark">
<td class="time" id="i_517437"><a href="/evergreen/2022-12-09#i_517437">13:44</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">mschell joined #evergreen</td>
</tr>
<tr id="id_l119" class="cont special dark">
<td class="time" id="i_517438"><a href="/evergreen/2022-12-09#i_517438">14:01</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">jvwoolf joined #evergreen</td>
</tr>
<tr id="id_l120" class="cont special dark">
<td class="time" id="i_517439"><a href="/evergreen/2022-12-09#i_517439">14:24</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">kworstell-isl joined #evergreen</td>
</tr>
<tr id="id_l121" class="new nick nick_mschell">
<td class="time" id="i_517440"><a href="/evergreen/2022-12-09#i_517440">14:37</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">Hi All, I'm hoping someone might have some insight into a problem I'm working on. If this isn't the proper forum just let me know (it's my first time on IRC).</td>
</tr>
<tr id="id_l122" class="cont nick nick_mschell">
<td class="time" id="i_517441"><a href="/evergreen/2022-12-09#i_517441">14:37</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">I'm trying to get the OPAC author display to show 'FirstName LastName' instead of 'LastName, FirstName'. There is a variable attr.author which pulls from MARC 100, which is what is displaying now but I wanted to see if using the MARC 245 $c would give me mostly what I want. The detailed view of the title in the OPAC (attrs.title.extended) shows</td>
</tr>
<tr id="id_l123" class="cont nick nick_mschell">
<td class="time" id="i_517442"><a href="/evergreen/2022-12-09#i_517442">14:37</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">the 245 $a and $c so I could maybe get it from there but how to only show the $c info and how would it affect the search link that is generated.</td>
</tr>
<tr id="id_l124" class="cont nick nick_mschell">
<td class="time" id="i_517443"><a href="/evergreen/2022-12-09#i_517443">14:37</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">Does this approach make sense? Is there a better way?</td>
</tr>
<tr id="id_l125" class="new nick nick_Bmagic dark">
<td class="time" id="i_517444"><a href="/evergreen/2022-12-09#i_517444">14:52</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">mschell: you might be able to get away with a little regular expression in the OPAC to tease out the pieces that you need</td>
</tr>
<tr id="id_l126" class="new nick nick_Dyrcona">
<td class="time" id="i_517445"><a href="/evergreen/2022-12-09#i_517445">15:02</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker | csharp_ : bug 1999274 I didn't link miker's branch because I thought you might like to add the bug # to the commit message.</td>
</tr>
<tr id="id_l127" class="new nick nick_pinesol dark">
<td class="time" id="i_517446"><a href="/evergreen/2022-12-09#i_517446">15:02</a></td>
<td style="color: #4e00a3" class="nick">pinesol</td>
<td class="msg ">Launchpad bug 1999274 in Evergreen "Performance of Search on PostgreSQL Versions 12+" [Medium,Confirmed] <a href="https://launchpad.net/bugs/1999274" title="https://launchpad.net/bugs/1999274">https://launchpad.net/bugs/1999274</a></td>
</tr>
<tr id="id_l128" class="new nick nick_Dyrcona">
<td class="time" id="i_517447"><a href="/evergreen/2022-12-09#i_517447">15:08</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">mschell: You could use the 245 $c if you prefer, just alter the XPath expression. I'd like to add that some think that using MARC in the OPAC is a mistake and a future version of the OPAC may switch to using display fields.</td>
</tr>
<tr id="id_l129" class="new special dark">
<td class="time" id="i_517448"><a href="/evergreen/2022-12-09#i_517448">15:08</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">sleary joined #evergreen</td>
</tr>
<tr id="id_l130" class="new nick nick_mschell">
<td class="time" id="i_517449"><a href="/evergreen/2022-12-09#i_517449">15:16</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">Bmagic, thanks I'll try that.</td>
</tr>
<tr id="id_l131" class="cont nick nick_mschell">
<td class="time" id="i_517450"><a href="/evergreen/2022-12-09#i_517450">15:17</a></td>
<td style="color: #00448c" class="nick">mschell</td>
<td class="msg ">Dyrcona, that is a future version I look forward too :)</td>
</tr>
<tr id="id_l132" class="new nick nick_Dyrcona dark">
<td class="time" id="i_517451"><a href="/evergreen/2022-12-09#i_517451">15:18</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">mschell: I wouldn't hold my breath. :)</td>
</tr>
<tr id="id_l133" class="new nick nick_Stompro">
<td class="time" id="i_517452"><a href="/evergreen/2022-12-09#i_517452">15:19</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">Is the eg2 splash page configurable somehow? It seems to not just be a template I can edit?</td>
</tr>
<tr id="id_l134" class="cont nick nick_Stompro">
<td class="time" id="i_517453"><a href="/evergreen/2022-12-09#i_517453">15:20</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">Ohh, staff portal page. see it now.</td>
</tr>
<tr id="id_l135" class="new nick nick_Bmagic dark">
<td class="time" id="i_517454"><a href="/evergreen/2022-12-09#i_517454">15:20</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">Stompro: I don't think so.</td>
</tr>
<tr id="id_l136" class="cont nick nick_Bmagic dark">
<td class="time" id="i_517455"><a href="/evergreen/2022-12-09#i_517455">15:20</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">you'd need to edit the source, then run the npm build steps</td>
</tr>
<tr id="id_l137" class="new nick nick_jihpringle">
<td class="time" id="i_517456"><a href="/evergreen/2022-12-09#i_517456">15:21</a></td>
<td style="color: #006b79" class="nick">jihpringle</td>
<td class="msg ">Bmagic: depends on the version, there's an admin interface starting in 3.8 or 3.9 I think (we definitely have it in 3.9)</td>
</tr>
<tr id="id_l138" class="new nick nick_Bmagic dark">
<td class="time" id="i_517457"><a href="/evergreen/2022-12-09#i_517457">15:21</a></td>
<td style="color: #9d6200" class="nick">Bmagic</td>
<td class="msg ">jihpringle++</td>
</tr>
<tr id="id_l139" class="new nick nick_mmorgan">
<td class="time" id="i_517458"><a href="/evergreen/2022-12-09#i_517458">15:22</a></td>
<td style="color: #00b65f" class="nick">mmorgan</td>
<td class="msg ">It's new in 3.9</td>
</tr>
<tr id="id_l140" class="new nick nick_Stompro dark">
<td class="time" id="i_517459"><a href="/evergreen/2022-12-09#i_517459">15:22</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">I testing on 3.9, so I get to try that out.</td>
</tr>
<tr id="id_l141" class="new nick nick_jeffdavis">
<td class="time" id="i_517460"><a href="/evergreen/2022-12-09#i_517460">15:25</a></td>
<td style="color: #c40092" class="nick">jeffdavis</td>
<td class="msg ">In addition to the new staff portal config options, we add an iframe to the page so we can display news updates from another source: <a href="http://git.sitka.bclibraries.ca/gitweb/?p=sitka/evergreen.git;a=commitdiff;h=45bfec6b5b1a7a8d0dfd898aa68926d17bb8f1ca" title="http://git.sitka.bclibraries.ca/gitweb/?p=sitka/evergreen.git;a=commitdiff;h=45bfec6b5b1a7a8d0dfd898aa68926d17bb8f1ca">http://git.sitka.bclibraries.ca/gitweb/?p=sitka/evergreen.git;a=commitdiff;h=45bfec6b5b1a7a8d0dfd898aa68926d17bb8f1ca</a></td>
</tr>
<tr id="id_l142" class="cont nick nick_jeffdavis">
<td class="time" id="i_517461"><a href="/evergreen/2022-12-09#i_517461">15:25</a></td>
<td style="color: #c40092" class="nick">jeffdavis</td>
<td class="msg ">I should probably turn that into a feature request.</td>
</tr>
<tr id="id_l143" class="new nick nick_Stompro dark">
<td class="time" id="i_517462"><a href="/evergreen/2022-12-09#i_517462">15:28</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">jeffdavis, thanks, that is interesting.</td>
</tr>
<tr id="id_l144" class="new nick nick_Dyrcona">
<td class="time" id="i_517463"><a href="/evergreen/2022-12-09#i_517463">15:28</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">So, we've got the storage listener running on our utility server, but no drones. It happened while the fine generator was running. I've seen something like this before, but IIRC I think the listener was what got shut down that time. Does this sound like a case of the listener somehow crashing and turning into a drone?</td>
</tr>
<tr id="id_l145" class="cont nick nick_Dyrcona">
<td class="time" id="i_517464"><a href="/evergreen/2022-12-09#i_517464">15:29</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">Nifty! The listener is immortal.</td>
</tr>
<tr id="id_l146" class="cont nick nick_Dyrcona">
<td class="time" id="i_517465"><a href="/evergreen/2022-12-09#i_517465">15:30</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">kill -9 works.</td>
</tr>
<tr id="id_l147" class="new special dark">
<td class="time" id="i_517466"><a href="/evergreen/2022-12-09#i_517466">15:31</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">kworstell-isl joined #evergreen</td>
</tr>
<tr id="id_l148" class="new nick nick_Stompro">
<td class="time" id="i_517467"><a href="/evergreen/2022-12-09#i_517467">15:34</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">Shoot, "it won’t try to merge branch, system, and consortial-level entries" I cannot have System portal with the odd entry for a specific branch. But I get that would be more complex to figure out.</td>
</tr>
<tr id="id_l149" class="new nick nick_jihpringle dark">
<td class="time" id="i_517468"><a href="/evergreen/2022-12-09#i_517468">15:53</a></td>
<td style="color: #006b79" class="nick">jihpringle</td>
<td class="msg ">ya, if you want different portals for different org units with slight differences you have to re-create the portal for each org unit you want (but there is a clone button)</td>
</tr>
<tr id="id_l150" class="new nick nick_miker">
<td class="time" id="i_517469"><a href="/evergreen/2022-12-09#i_517469">15:58</a></td>
<td style="color: #14000e" class="nick">miker</td>
<td class="msg ">Dyrcona: I'll clean up the commit message and push another branch. if you're in a position to test the patch in both opac and staff search, it'd be a much appreciated cross check.</td>
</tr>
<tr id="id_l151" class="new nick nick_Dyrcona dark">
<td class="time" id="i_517470"><a href="/evergreen/2022-12-09#i_517470">16:00</a></td>
<td style="color: #7d008d" class="nick">Dyrcona</td>
<td class="msg ">miker: Sure thing!</td>
</tr>
<tr id="id_l152" class="new special">
<td class="time" id="i_517471"><a href="/evergreen/2022-12-09#i_517471">16:59</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">mmorgan left #evergreen</td>
</tr>
<tr id="id_l153" class="new nick nick_Stompro dark">
<td class="time" id="i_517472"><a href="/evergreen/2022-12-09#i_517472">17:00</a></td>
<td style="color: #00acc7" class="nick">Stompro</td>
<td class="msg ">Who do I contact to add something to the tabular release notes?</td>
</tr>
<tr id="id_l154" class="new nick nick_jihpringle">
<td class="time" id="i_517473"><a href="/evergreen/2022-12-09#i_517473">17:19</a></td>
<td style="color: #006b79" class="nick">jihpringle</td>
<td class="msg ">Stompro I believe jweston</td>
</tr>
<tr id="id_l155" class="new special dark">
<td class="time" id="i_517474"><a href="/evergreen/2022-12-09#i_517474">17:37</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">stompro_home joined #evergreen</td>
</tr>
<tr id="id_l156" class="cont special dark">
<td class="time" id="i_517475"><a href="/evergreen/2022-12-09#i_517475">17:38</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">stompro__ joined #evergreen</td>
</tr>
<tr id="id_l157" class="cont special dark">
<td class="time" id="i_517476"><a href="/evergreen/2022-12-09#i_517476">18:45</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">jonadab joined #evergreen</td>
</tr>
<tr id="id_l158" class="cont special dark">
<td class="time" id="i_517477"><a href="/evergreen/2022-12-09#i_517477">19:58</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">jvwoolf left #evergreen</td>
</tr>
<tr id="id_l159" class="cont special dark">
<td class="time" id="i_517478"><a href="/evergreen/2022-12-09#i_517478">20:37</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">stompro__ joined #evergreen</td>
</tr>
<tr id="id_l160" class="cont special dark">
<td class="time" id="i_517479"><a href="/evergreen/2022-12-09#i_517479">20:38</a></td>
<td style="color: #ad7900" class="nick"></td>
<td class="msg ">stompro_home joined #evergreen</td>
</tr>
</table>
</form>
<span id="bottom" />
<p>
<a href="/evergreen/2022-12-08" rel="prev">
← Previous day</a> |
<a href="/">Channels</a> |
<a href="/evergreen/">#evergreen index</a> |
<a href="/evergreen/today">Today</a> |
<a href="/evergreen/2022-12-10" rel="next">
Next day →</a> |
<a href="/evergreen/search">Search</a> |
<a href="http://www.google.com/search?q=site%3Airc.evergreen-ils.org+inurl%3Aevergreen">Google Search</a>
| <a href="/evergreen/2022-12-09/text">Plain-Text</a> |
<a href="/evergreen/2022-12-09/summary">summary</a>
| <a href="https://webchat.freenode.net/?channels=evergreen">Join Webchat</a>
</p>
<div id="footer">
<p>Powered by <a href="http://moritz.faui2k3.org/en/ilbot">ilbot</a>.
<strong>Any questions? Ask bshum in #evergreen on irc.libera.chat</strong></p>
<p>
Evergreen is open source software, freely
licensed under <a href="http://www.gnu.org/licenses/gpl-2.0.html"
target="_blank">GNU GPLv2</a> or later.<br /> The <a
href="http://evergreen-ils.org/blog/?p=583" target="_blank">Evergreen Project is
a member</a> of <a href="http://sfconservancy.org/" target="_blank">Software
Freedom Conservancy</a>.
</p>
</div>
<script type="text/javascript" src="/s/jquery.min.js"></script>
<script type="text/javascript">
IlbotConfig = {
base_url: '/',
channel: 'evergreen',
day: '2022-12-09',
still_today: false
};
</script>
<script type="text/javascript" src="/s/dynamic.js"></script>
</body>
</html>
<!-- -->