Time
S
Nick
Message
07:53 jri joined #dataverse
07:55 jri joined #dataverse
10:27 jri joined #dataverse
12:41 jri joined #dataverse
12:44 jri joined #dataverse
14:35 rebecabarros joined #dataverse
14:38 rebecabarros
pdurbin: good morning =). I've started the doc about some DCM installation notes. Here it is: https://docs.google.com/document/d/1m5JZnLNhsvwixaV16eyP9o0JMO7tAd5I1tIDcb3_g9g/edit?usp=sharing
14:44 rebecabarros
not sure If will be helpful but i've listed some of the things that i had to do/change during the installation.
14:47 rebecabarros joined #dataverse
14:48 rebecabarros
pameyer: I've noticed today that running dcm-test01 from Dataverse server I've got the json file created in /requests. It has the following content: {"datasetId":"23", "userId":"2","datasetIdentifier":"23"}. is that right? But dcm-test02 (and sr.py) still gives me 404 error.
14:52 rebecabarros joined #dataverse
15:01 donsizemore joined #dataverse
16:15 pameyer joined #dataverse
16:16 pameyer
rebecabarros: what interval do you have for the cron job (post_upload)?
16:23 rebecabarros joined #dataverse
16:23 rebecabarros
pameyer: what do you mean? I did not understand
16:29 pameyer joined #dataverse
16:30 pameyer
404 means that there's no script (yet). so either there's a problem with job doing the setup and script creation, or it's not getting run
16:31 Thalia_UM joined #dataverse
16:31 Thalia_UM
Good morning :)
16:31 Thalia_UM
Philip I have a question about dataverse installation
16:31 rebecabarros
pameyer: how/where I check the interval value?
16:34 rebecabarros88 joined #dataverse
16:34 pameyer
`crontab -l` , or check `/etc/cron*`
16:34 rebecabarros88 left #dataverse
16:38 rebecabarros joined #dataverse
16:40 rebecabarros
pameyer: should I run with the specific user dcmu? or my personal user it's ok?
16:40 rebecabarros
cause my personal username gives me: no crontab
16:42 pameyer
it needs to create os accounts, so probably either root's crontab for sudo config for dcmu
16:43 pdurbin
rebecabarros: these notes you made are great. Thanks so much!
16:58 pameyer joined #dataverse
17:03 rebecabarros joined #dataverse
17:05 rebecabarros
pameyer: I do have a user (dcmu) which is setted in dcm/vars/main.yml
17:06 rebecabarros
either way, there is 'no crontab' for any user, even for root
17:08 pdurbin
Thalia_UM: good morning. What's your question? Sorry for any cross-talk, rebecabarros
17:16 Thalia_UM
don't worry Philip
17:16 Thalia_UM
Well
17:20 Thalia_UM
When entering dataverse, I have it installed in local, I see the page of glassfish server and 404 error for the pages of dataverse and that has made that I have to be installing every time I see the error. Is there another way to recover the pages?
17:20 pdurbin
The community call is going on right now, for anyone who's interested: https://dataverse.org/community-calls
17:20 Thalia_UM
The error is HTTP Status 404- Not found
17:20 Thalia_UM
Description: The requested resource is not available
17:29 jri joined #dataverse
17:30 Thalia_UM
How can I ask or what do I do in the document?
17:33 Thalia_UM
Thank you Philip
17:33 Thalia_UM
You are great
17:33 Thalia_UM
:)
17:38 pameyer
Thalia_UM: this sounds to me like glassfish didn't stop cleanly at some point
17:38 pameyer
I'd usually stop glassfish; then do `ps aux | grep glassfish` and kill any remaining jobs, and restart
17:38 pameyer
it might not solve your problem, but it might be worth a shot
17:39 Thalia_UM
What does that mean?
17:39 pdurbin
Thalia_UM: can you please post a screenshot at https://imgur.com or similar?
17:40 Thalia_UM
Sure
17:42 Thalia_UM
https://imgur.com/yi7DxKW
18:01 rebecabarros joined #dataverse
18:23 pdurbin
Ok, so you're getting a Glassfish 404 error at http://localhost:8080/dataverse.xhtml . This can mean that the Dataverse war file isn't deployed at all.
18:23 pdurbin
Thalia_UM: what does `asadmin list-applications` show?
18:28 Thalia_UM
dataverse <ejb, web>
18:28 Thalia_UM
Command list-applications executed successfully.
18:28 Thalia_UM
That
18:29 pdurbin
Interesting. What do you see at http://localhost:8080 ?
18:30 Thalia_UM
https://imgur.com/jh1CkJ8
18:31 pdurbin
Ok, 404 there too. Has src/main/webapp/WEB-INF/glassfish-web.xml been change? Especially "context-root"?
18:32 Thalia_UM
No
18:33 Thalia_UM
Why?
18:33 pdurbin
Sometimes Netbeans makes change to that file. Please see http://guides.dataverse.org/en/4.8.3/developers/dev-environment.html#context-root-in-glassfish-web-xml-munged-by-netbeans
18:37 Thalia_UM
But I don't use netbeans, I downloaded the installer and follow the steps in the installation guide of dataverse.
18:38 pdurbin
Ok. Where did you download the war file from?
18:39 pameyer joined #dataverse
18:42 Thalia_UM
https://github.com/IQSS/dataverse/releases/tag/v4.7
18:43 Thalia_UM
It installs well but the next day that error appears.
18:44 pdurbin
Very strange. Have you tried undeploying the war file, restarting Glassfish, and reploying the war file?
18:44 Thalia_UM
Yes
18:45 Thalia_UM
5 times
18:45 pameyer
rebecabarros: I'm not sure which internal value you were asking about. The ansible role doesn't set up the cron entry; this is definately in the docs for the original system, but I haven't verified that it's in the docs in the DCM repo
18:46 Thalia_UM
That problem was barely generated last week, that's how I was all week.
18:46 pameyer
pdurbin: do you remember if the timer-related changes for harvesting were in v4.7? I'm wondering if that could be related
18:47 pdurbin
pameyer: maybe but I thought those timer problems resulting in users being unable to deploy the war file at all. Maybe I'm misremembering.
18:48 pdurbin
If the war file is deployed there should be some signs of life.
18:48 pameyer
yeah - that's what made me think killing glassfish entirely and restarting might help
18:49 pdurbin
Thalia_UM: sometimes I use the "version" API to see if the war file has been deployed. The output looks like this: https://demo.dataverse.org/api/info/version
18:49 pameyer
not sure if that would be a factor if glassfish wasn't stopped but the application disappeared
18:49 pdurbin
Thalia_UM: do you get any JSON output at http://localhost:8080/api/info/version ?
18:51 rebecabarros
pameyer: the one that you mentioned here: http://irclog.iq.harvard.edu/dataverse/2017-11-21#i_60482
18:54 Thalia_UM
Nothing, Only Http Status 404
18:55 pameyer
rebecabarros: something like `*/5 * * * * /usr/local/dcm/scn/post_upload.bash >> /var/log/post_upload.log` from `crontab -l` is the interval I was referring to (aka - once every 5 minutes)
18:55 pameyer
does that make sense?
18:55 pdurbin
Thalia_UM: ok. Thanks. Any more ideas, pameyer ? Thalia_UM I'm close to telling you that you should email support
18:55 pameyer
also; irc logs are helpful
18:56 pameyer
pdurbin, Thalia_UM: I'm still thinking cold-stopping glassfish is worth a look
18:57 pameyer
even if I'm wrong, I'm still thinking it
18:57 pdurbin
pameyer: I mention IRC logs in my article about transparency: https://groups.google.com/d/msg/dataverse-community/brxCn1E9tX0/VbsNz4u8BgAJ :)
18:57 pameyer
transparency, and folks with poor memory
18:57 pameyer
like me
18:57 pdurbin
EXACTLY, me too :)
18:58 pdurbin
Thalia_UM: if it helps, there are some extra cleanup steps you can take such as removing the "generated" directory after you've stopped Glassfish. Please see http://guides.dataverse.org/en/4.7/admin/troubleshooting.html
18:59 Thalia_UM
I have also applied that
19:01 pdurbin
All I can think to do is to install a fresh copy of Glassfish. But that means you would need to run the installer again.
19:01 rebecabarros
pameyer: it does, but I don't have return like this from 'crontab -l', and my "/etc/crontab" file is empty, without any entry.
19:03 pameyer
rebecabarros: if you run post_upload.bash directly (as root), that should generate a script and transfer user (if all goes well). if so, `crontab -e` should let you setup that job to run periodicly
19:03 pameyer
if not; then we'll have to do more troubleshooting
19:05 rebecabarros
pameyer: so, should I try to run post_upload.bash and after that try again to run dcm-test-02 (sr.py), right? Let me try that
19:07 pameyer joined #dataverse
19:07 pameyer
rebecabarros: right
19:07 pameyer
you could check $DEPOSIT prior to dcm-test02; the request file should move and there should be a home directory for the transfer user
19:09 rebecabarros
pameyer: I see here that in post_upload I have to set manually DEPOSIT, HOLD AND SRC. $DVAPIKEY, $DVHOST and $DOI_SHOULDER do not need to be setted, right? meaning some other script will send this values?
19:13 pameyer
$DVAPIKEY,$DVHOST and $DOI_SHOULDER should be environmental variables; I believe ansible should put them into root's ~/.bashrc
19:21 rebecabarros
pameyer: that's right. I ran post_upload and went ok, got 'post_upload completed at Tue Nov 21 16:16:46 -03 2017'
19:29 pameyer
(doh) I'm just realizing partly why multi-tasking is bad. In other words, post_upload isn't (yet) the problem.
19:29 pameyer
is there anything in /var/log/rq_worker.log?
19:35 rebecabarros
pameyer: https://pastebin.com/DfvRe0RE - I've ran dcm-test01 (and the file was created in /requests), than I ran dcm-test02 using the same dataset number. That' the content of rq_worker
19:36 rebecabarros
the file '/usr/local/dcmu/dcm/deposit/requests/4652.json' exists
19:38 pameyer
is rq running as root, or another user?
19:39 rebecabarros
how can I check?
19:42 pameyer
`ps aux | grep rq` should tell you
19:44 rebecabarros
Its running as root
19:47 pameyer
am I remembering correctly that selinux is off?
19:48 rebecabarros
Yes. It's set as permissive
19:50 pameyer
could you post the permissions for /usr/local/dcmu/dcm/deposit/requests/4652.json?
19:53 rebecabarros
pameyer: -rw-r--r--. 1 lighttpd lighttpd 57 Nov 21 16:23 4652.json
20:09 pameyer joined #dataverse
20:17 rebecabarros left #dataverse
20:30 pameyer joined #dataverse
21:13 pameyer joined #dataverse
21:42 pameyer joined #dataverse
21:44 pameyer
rebecabarros: I'm not able to reproduce that problem; but I'd suggest restarting rq and confirming that 4652.json is actually readable by root (by permissions, it should be)
22:01 jri joined #dataverse
23:27 Thalia_UM left #dataverse
