Time
S
Nick
Message
03:38 jri joined #dataverse
03:51 hello joined #dataverse
03:52 hello8523 joined #dataverse
04:46 icarito[m] joined #dataverse
05:23 rigelk joined #dataverse
06:38 jri joined #dataverse
06:54 juancorr
Morning @pdurbin, I was at home before the meeting. I am sorry but I have not enough knowledge about application servers and we have not special requirements to new Dataverse versions related to them.
08:01 cryptoclidus joined #dataverse
08:02 cryptoclidus
Hi there. Just checking in - I've made a Jupyter notebook cataloging the top commenters for any given month on this chat
08:53 pdurbin
juancorr: no problem, thanks
08:53 pdurbin
cryptoclidus: this is great!!! Thanks!
08:54 cryptoclidus
I put up a pull request!
08:54 pdurbin
I'll go merge it. One sec.
08:56 cryptoclidus21 joined #dataverse
08:56 pdurbin
merged!
08:56 cryptoclidus21
cool!
08:56 cryptoclidus21
For the monthly superusers graph, there's one line that says month = "April"
08:57 cryptoclidus21
if you just change the month, and update all the code under it, you'll get an updated graph for that month
08:57 cryptoclidus21
You have to spell the month correctly though (full name, capital first letter)
08:57 pdurbin
sounds easy enough :)
08:58 pdurbin
Ok, and the other one is "all time". This is so interesting.
09:00 cryptoclidus21
you definitely chat the most!
09:00 pdurbin
Heh, I'm not surprised. :)
09:00 cryptoclidus21
Also I noticed that at least here, there's an issue when the same person enters with slightly different nicknames
09:00 pdurbin
Right, yes, this happens a lot. :(
09:00 cryptoclidus21
so there's some 'duplicates' like pdurbin, *pdurbin and some others
09:00 cryptoclidus21
and like mine now, I don't know how the 21 got there
09:01 pdurbin
well, *pdurbin is when you "emote" with /me if you're familiar with that
09:01 * pdurbin
dances
09:01 cryptoclidus21
I guess in other chatrooms with fixed nicknames it's not an issue
09:01 cryptoclidus21
OH hah that's hilarious
09:01 cryptoclidus21
that means you've emoted a lot of times.
09:02 * pdurbin
dances and sings
09:02 cryptoclidus21
so you could technically filter top emoters too
09:02 cryptoclidus21
anyways it's past bedtime here
09:02 cryptoclidus21
I better check out
09:02 cryptoclidus21
ttyl!
09:02 pdurbin
Thanks again! Awesome stuff!
09:03 cryptoclidus21
Thanks! Definitely would be happy for any feedback.
09:03 pdurbin
cool
11:01 stefankasberger joined #dataverse
11:03 stefankasberger
@pdurbin: coming back in regards to the offer to run pyDataverse on your Jenkins instance. Maybe we could talk it through during my stay at harvard. or maybe even implement it. what do you think?
11:09 pdurbin
stefankasberger: sure! That sounds fine.
11:10 pdurbin
And speaking of pyDataverse, yesterday I tweeted about it: https://twitter.com/philipdurbin/status/1133335435340636161 :)
11:15 stefankasberger
great, thanks.
11:21 pdurbin
stefankasberger: is there anything you need? Any blockers? Or anything you want to help with? Maybe you're busy writing your talk. I should be writing my talk. :)
11:34 poikilotherm joined #dataverse
11:39 poikilotherm
Hey guys :-)
11:40 pdurbin
poikilotherm: hey, I have a new jupyter notebook to show you. :)
11:41 poikilotherm
Sounds nice :-)
11:41 pdurbin
https://github.com/pdurbin/dataverse-irc-metrics/pull/1
11:41 pdurbin
Soon you'll surpass pameyer. :)
11:51 poikilotherm
LOL
11:51 poikilotherm
Just about 405 messages :-D
11:51 poikilotherm
404
11:51 poikilotherm
403
11:51 poikilotherm
402
11:51 poikilotherm
:-D
11:51 pdurbin
heh
11:52 pdurbin
I was thinking... we *could* count characters instead of lines. But whatever. :)
11:52 stefankasberger
i am really busy right now, yes. am implementing the metadata data models for dataverses and datasets, to do easy import/export.
11:52 stefankasberger
and write the presentation
11:53 poikilotherm
pdurbin I really appreciate that you moved on during tech hours and your write up that triggered movement
11:53 pdurbin
stefankasberger: cool, you don't sound blocked. You're on a roll. :)
11:53 pdurbin
poikilotherm: during tech hours I tried to channel you. I wore a fake goatee.
11:54 poikilotherm
ROFLMAO
11:54 pdurbin
I think writing this stuff up is important. Otherwise, how can anyone know what we talked about.
11:56 pdurbin
I just tried spinning up an ec2 instance with this: zipurl: https://s3-eu-west-1.amazonaws.com/payara.fish/Payara+Downloads/5.192/payara-5.192.zip
11:56 pdurbin
but I'm having ansible trouble :(
11:57 pdurbin
I don't know if you've already parameterized the zip url like that in your stuff.
12:00 dataverse-user joined #dataverse
12:03 poikilotherm
Kind of...
12:03 poikilotherm
https://github.com/IQSS/dataverse-kubernetes/blob/98be266b3c01423526accab8f5f07cbaed2ca7d0/docker/dataverse-k8s/Dockerfile#L21
12:03 poikilotherm
You could override that.
12:04 poikilotherm
My plan was to swap the base image from centos:7 to payara:5.192
12:04 poikilotherm
https://hub.docker.com/r/payara/server-full
12:04 poikilotherm
Oh I just saw that 5.192 has been tagged an hour ago...
12:05 poikilotherm
Coincedence? Maybe not :-D
12:05 pdurbin
Sure, it makes sense to eventually switch the base image to Payara for your stuff.
12:05 poikilotherm
https://github.com/payara/docker-payaraserver-full/pull/89
12:05 poikilotherm
Hehehe
12:06 poikilotherm
I think this has been merged on purpose :-D
12:07 pdurbin
:)
12:08 poikilotherm
Hmm I think I should give this a quick shot :-D
12:08 poikilotherm
But maybe better work on the deployment issues first...
12:09 pdurbin
I'm hacking on https://github.com/IQSS/dataverse-metrics
12:11 pdurbin
stefankasberger: when I'm done I'm hoping to try pyDataverse and work on https://github.com/IQSS/dataverse/issues/5235 (if nothing else becomes a priority first)
12:12 donsizemore joined #dataverse
12:38 pdurbin
donsizemore: good morning!
12:41 pdurbin
https://rin.lipi.go.id was just upgraded. Can you please add it to https://dataversemetrics.odum.unc.edu/dataverse-metrics/config.json ?
12:46 donsizemore
@pdurbin das' a dataverse? it purty
12:48 donsizemore
@pdurbin added and updating. want me to add it to config.json.sample in the repo?
13:13 pdurbin
donsizemore: thanks! That's ok, I'm in that code anyway so I'll get it. I was thinking, maybe I should back up the live config to https://github.com/IQSS/dataverse.org to something.
13:17 Richard_Valdivia joined #dataverse
13:19 pdurbin
It's funning, now that "LIPI Dataverse" appears at https://dataverse.org/metrics there are some scrollbars. I'll go make the iframe taller.
13:23 pdurbin
funny*
13:24 pdurbin
hmm, but making it taller doesn't help. ¯\_(ツ)_/¯
13:28 poikilotherm
No scrollbars here...
13:29 pdurbin
This is on Firefox. And I *think* I fixed it but there's some caching. Fingers crossed.
13:30 poikilotherm
This is Firefox 60.6esr on Linux. No scrollbars here even when deleting local cache.
13:31 pdurbin
I'm talking about server side caching but I'm glad you don't see the scrollbars. Hopefully I'm the only one who sees them. :)
13:31 donsizemore
@pdurbin sumbody threw a 404
13:32 pdurbin
better error handling needed?
13:32 donsizemore
it currently doesn't handle HTTP error codes
13:33 pdurbin
yeah :(
13:33 pdurbin
you'd take any error handling :)
13:34 pdurbin
what should happen when there's a non 200 HTTP code? do you want a log or something?
13:37 donsizemore
@pdurbin stderr is fine by me. i'd just like for it to move on to the next host
13:37 donsizemore
(sorry, was vacuuming my office)
13:38 pdurbin
ok, let me look into stderr, thanks
13:40 pdurbin
I'll try to roll it into this new metric for github contributors.
14:14 donsizemore
@pdurbin it's https://rin.lipi.go.id throwing the 404
14:32 pdurbin
d'oh!
14:39 pdurbin
I see. nginx 404 at https://rin.lipi.go.id/api/info/metrics/datasets/bySubject for example. Hmm.
14:39 donsizemore
@pdurbin should in fact be https://data.lipi.go.id/
14:39 donsizemore
@pdurbin I knew that home page was too faincy
14:39 pdurbin
oh!
14:39 pdurbin
good catch
14:40 donsizemore
today was cleaning morning, so i've been distracted
14:45 donsizemore
@pdurbin so, i was adding a jacoco switch to dataverse-ansible, per the dev guide
14:46 donsizemore
@pdurbin it instruments the warfile, then attempts to deploy: Exception while deploying the app [dataverse] : org/jacoco/agent/rt/internal_c13123e/Offline. Related annotation information: annotation [@javax.ejb.TransactionAttribute(value=REQUIRES_NEW)] on annotated element [public void edu.harvard.iq.dataverse.GuestbookResponseServiceBean.save(edu.harvard.iq.dataverse.GuestbookResponse)] of type [METHOD]
14:46 donsizemore
curious if you've seen that before?
14:54 pdurbin
let me search my email
14:55 pdurbin
donsizemore: I don't believe so.
14:56 pdurbin
But none of us ever tried Pete's write up, unfortunately.
14:57 donsizemore
well, the switch is soon to be available in dataverse-ansible =) wanted it for jenkins.dataverse
14:57 pdurbin
Can we back up to what the goal is? The goal is still to replicate the "Test Result Trend" graph at https://build.hmdc.harvard.edu:8443/job/phoenix.dataverse.org-apitest-develop/ right?
15:05 donsizemore
@pdurbin that's one of them... i was just going to add jacoco as a switch, which seemed easy enough.
15:06 pdurbin
ok, and I think I have another installation to add to the config :)
15:06 donsizemore
which one?
15:07 pdurbin
donsizemore: only one dataset so far but if you could add it, it would be most appreciated: http://research-data.urosario.edu.co
15:07 pdurbin
"Datos Brutos" is the name of the dataset
15:07 donsizemore
done. lemme let this run complete and i'll submit a PR to update the sample config.json
15:08 pdurbin
sounds good! thanks!
15:10 pdurbin
donsizemore: you might need to run https://github.com/IQSS/dataverse-metrics/blob/master/update-all-installations-list.sh too
15:23 donsizemore
also done
15:29 pdurbin
cool but there's no LIPI
15:29 pdurbin
so maybe there's a mismatch of hostnames or something?
15:32 pdurbin
yeah, https://rin.lipi.go.id appears in all-dataverse-installations.json for the map but like you said it's not the "real" hostname, which is https://data.lipi.go.id
15:38 poikilotherm
Oi donsizemore... https://github.com/IQSS/dataverse-kubernetes/issues/60#issuecomment-496989761
15:50 pdurbin
bricas_: how dare you have a lower github id than me
16:16 poikilotherm
donsizemore I think I found a bug in Minikube :-(
18:37 donsizemore
@pdurbin prep_IT.bash completed successfully!
18:47 pdurbin
donsizemore: nice!
18:48 pdurbin
Also, I'm working on getting those two hostnames aligned.
18:56 donsizemore
@pdurbin me API test suite test runs still die in the same place: [ERROR] testPrivateUrl Time elapsed: 316.773 s <<< FAILURE! junit.framework.AssertionFailedError: expected:<200> but was:<500> at edu.harvard.iq.dataverse.api.DatasetsIT.testPrivateUrl(DatasetsIT.java:948)
18:58 donsizemore
@pdurbin which may have me at parity with #5826 IDK
19:00 pdurbin
you might be running into https://github.com/IQSS/dataverse/issues/5827
19:00 pdurbin
brittle tests
19:00 pdurbin
but they were passing on phoenix the last time I ran them
19:03 donsizemore
I seem to reproduce it reliably
19:06 pdurbin
donsizemore: I just put out a new release with your updated sample file and an updated "all installations" file: https://github.com/IQSS/dataverse-metrics/releases/tag/v0.2.4
19:07 pdurbin
if you put the new all installations file into place the lipi installation should appear in the list at the bottom (the hostnames will now match)
19:07 donsizemore
want me to upgrade dataversemetrics?
19:07 donsizemore
(or were the two new installations the only change)
19:08 pdurbin
well, I'm trying to fix a bug
19:08 pdurbin
the installations only appear at the bottom if the hostname matches
19:08 pdurbin
does that make sense?
19:09 pdurbin
here's where that logic is: https://github.com/IQSS/dataverse-metrics/blob/v0.2.4/plots.js#L286
19:11 pdurbin
basically, the code is iterating through all-dataverse-installations.json and it only spits out a <li> when it matches a hostname found in config.json
19:11 pdurbin
so I updated all-dataverse-installations.json to have the actual url of the dataverse installation for LIPI
19:21 pdurbin
(also, I'm working on running the api tests in docker-aio on my mac)
19:43 pdurbin
Hmm. DataversesIT.testMoveDataverse:364 JSON path data.total_count doesn't match. Expected: <1> Actual: 0 ... that's on 02c36c4c9
19:43 pdurbin
same as what's on phoenix where the tests pass
19:44 donsizemore
dataversemetrics is upgraded to v0.2.4 and updating now
19:45 pdurbin
donsizemore: thanks! And I see "LIPI Dataverse" on the list.
19:49 pdurbin
donsizemore: I ran the test again and got BUILD SUCCESS
19:49 donsizemore
!!!
19:49 pdurbin
¯\_(ツ)_/¯
19:49 pdurbin
conf/docker-aio/run-test-suite.sh I mean
19:50 donsizemore
no that's a happy excite. run-test-suite started throwing 500s for me again this afternoon
19:50 pdurbin
heh
19:50 pdurbin
should I spin up an ec2 and try it there?
19:50 pdurbin
or do you want to spin one up and add my ssh keys?
19:51 pdurbin
might be nice to troubleshoot from the same box
19:52 donsizemore
you're welcome to. i made dataverse-jenkins match phoenix, virtual-hardware-wise, for this reason
19:52 pdurbin
ok
19:53 pdurbin
so you're running docker-aio on the jenkins server itself?
20:23 poikilotherm joined #dataverse
