Time
S
Nick
Message
12:57 iqlogbot joined #dataverse
12:57 Topic for #dataverse is now Dataverse is open source research data repository software: http://dataverse.org | IRC Logs: http://irclog.iq.harvard.edu/dataverse/today | Who's who: https://docs.google.com/spreadsheets/d/16h3jv24usMGq18495C-JA-yNcQCKiKDa65MTraNDd7k/edit?usp=sharing
13:48 donsizemore joined #dataverse
13:59 pdurbin joined #dataverse
13:59 pdurbin
bricas: whoops. We should be talking in here.
14:00 pdurbin
You said you took a look at https://github.com/IQSS/dataverse/issues/4665 ?
14:25 cthao joined #dataverse
14:27 cthao
hi all, I am working on issue #4762 (scaling solr), does anyone know if solr Index replication is favorable over Sharding?
14:28 cthao
https://github.com/IQSS/dataverse/issues/50 is pretty outdated but mentioned it a bit
14:55 cthao joined #dataverse
14:56 pdurbin
cthao: I do see the word "sharding" in a URL in that issue but not much discussion.
14:57 cthao
My work so far is having a replication master/slave solr setup
14:57 pdurbin
https://cwiki.apache.org/confluence/display/solr/Distributed+Search+with+Index+Sharding now redirects to https://lucene.apache.org/solr/guide/6_6/distributed-search-with-index-sharding.html which says "It is highly recommended that you use SolrCloud when needing to scale up or scale out."
14:57 pdurbin
I'm not sure what the Solr 7 docs say.
14:59 cthao
yeah I know
14:59 cthao
I read that article but Dan doesn
14:59 cthao
doesn't recommend having zookeeper complicating the setup
15:00 pdurbin
ok
15:02 cthao joined #dataverse
15:03 pdurbin
cthao: there's a memory leak it http://chat.dataverse.org by the way: https://github.com/IQSS/chat.dataverse.org/issues/3
15:04 pdurbin
So you might want to try Adium or Pidgin: https://github.com/IQSS/chat.dataverse.org/issues/2
15:04 pdurbin
(I use weechat.)
15:04 pdurbin
leak in*
15:06 cthao
I would definitely take a look
15:06 cthao
sure
15:07 cthao joined #dataverse
15:07 cthao
I think I was kicked out for some reasons
15:09 pdurbin
It's probably the memory leak. Not your fault.
15:14 pdurbin
cthao: thanks for resolving the merge conflicts in https://github.com/IQSS/dataverse/issues/4763 . I just moved it back into QA and left a comment.
15:23 donsizemore joined #dataverse
15:32 cthao
No problem. it's mostly Tommy's work
16:26 pameyer joined #dataverse
18:02 pdurbin
pameyer: this can wait until you're over at our office but I'm wondering if you've seen the new todo list at https://github.com/IQSS/dataverse/issues/4703#issuecomment-406633874
18:03 pameyer
pdurbin: I did see it
18:04 pdurbin
Cool. When I hacked on DCM a bit I used Vagrant.
18:04 pdurbin
But it sounds like we're using Docker now.
18:04 pameyer
talked to mdunlap Wed-ish about possibly doing a pre code review; haven't done that yet
18:04 pdurbin
Which is fine.
18:05 pdurbin
In the Dataverse code base I'm happy to keep Vagrant working and I'm glad for your new docker-aio thing.
18:05 pameyer
they both work - I'm still using multi-VM vagrant for full-up integration testing. trying to stick with docker for non-fork / non-repository specific stuff due to overhead of multi-system vagrant
18:05 pdurbin
I'm curious how you feel about maintaining both Vagrant and Docker in the DCM code base.
18:06 pameyer
well, I can't run vagrant in jenkins
18:06 pameyer
but haven't given it much (or any) thought
18:08 pdurbin
Ok, Maybe we can touch base next time we're both at standup or something.
18:09 pameyer
pdurbin: do you know offhand where the ?persistentIdentifier=$foo gets into the dataset page?
18:10 pdurbin
pameyer: my money's on `<f:viewParam name="persistentId" value="#{DatasetPage.persistentId}"/>` at https://github.com/IQSS/dataverse/blob/v4.9.1/src/main/webapp/dataset.xhtml#L49
18:10 pdurbin
unless I misunderstand you
18:10 pdurbin
which has never happened in the history of history
18:10 pameyer
I'm sure it's happened at least once ;)
18:11 pameyer
I ran into an odd case where I could trigger 500s depending on what I passed there, and was poking around to see if there was a place to prevent that
18:12 pdurbin
sounds like a bug
18:12 pameyer
then my java skills and available time were inadaquate, so I went with apache
18:12 pdurbin
can you reproduce it on http://phoenix.dataverse.org ?
18:13 pameyer
maybe - let's find out
18:14 pameyer
yup
18:14 * pdurbin
stares at a giant stack trace
18:14 pdurbin
pameyer: do you feel like opening an issue for it?
18:15 pameyer
does this feel security related to you?
18:15 pameyer
seemed potentially that way to me, which is why there isn't one yet
18:16 pdurbin
no, but you have a better head for security than I do. Can I paste the PSQLException here in this PUBLICLY LOGGED channel? :)
18:16 * pdurbin
tosses iqlogbot a snack
18:17 pameyer
I've seen that exception trace ;)
18:17 pdurbin
good
18:20 pameyer
maybe another thing we should touch base on next time we're both at standup
18:22 pdurbin
pameyer: sure. Sometimes we create issues with obscure titles. So we can estimate it at least.
18:23 pameyer
pdurbin: I've done the obscure title trick before ...
18:26 pdurbin
pameyer: I can reproduce it on my laptop
18:28 pameyer
it's easy to reproduce :(
18:29 pdurbin
T-T
18:29 pameyer
definately a bug, and I'm likely being overly cautious about handling it
18:31 pdurbin
I'd say so but whatever. :)
18:51 pdurbin
pameyer: you can just add an extra try/catch if you want to fix it in your fork
18:53 pdurbin
er, an extra catch in the existing try
18:55 pameyer joined #dataverse
18:56 pameyer
pdurbin: that was my first 2 tries... blocking in at apache is easier
18:58 pdurbin
ok, whatever works :)
19:25 pameyer
pdurbin: I thought that test_006_ReplaceFileGood had been fixed in develop. am I remembering wrong, or is that just not merged yet?
19:26 pdurbin
lemme check
19:27 pdurbin
pameyer: https://github.com/IQSS/dataverse/issues/4866 is the issue and the pull request hasn't been merged yet: https://github.com/IQSS/dataverse/pull/4867
19:28 pameyer
gotcha - thanks
19:29 pdurbin
lotta stuff getting merged right now
19:29 * pdurbin
crosses fingers
19:30 pameyer
yeah - busy friday for kcondon
20:10 pdurbin
fingers still crossed
20:27 pdurbin
pameyer: merged and I kicked off the test. Why wait for "@nightly" ;)
20:29 pameyer
:)
20:35 pdurbin
waiting for https://build.hmdc.harvard.edu:8443/job/phoenix.dataverse.org-apitest-develop/ to finish
20:35 pdurbin
I'm on the edge of my seat. :)
20:43 pameyer
I know microwaves and teapots take longer when you watch them - do jenkins servers?
20:48 pdurbin
Done! Fixed! The API suite is passing again.
20:48 pdurbin
I think that means it's quittin time.
20:49 pameyer
nice!
20:49 pameyer
... to both
20:51 pdurbin
Have a good weekend, everyone.
20:51 pdurbin left #dataverse
21:55 pameyer left #dataverse
22:10 icarito[m] joined #dataverse
