Time
S
Nick
Message
00:16 pdurbin
hey there jls_ucb
03:12 jri joined #dataverse
07:01 jri joined #dataverse
07:04 jri joined #dataverse
12:19 donsizemore joined #dataverse
12:34 Frank joined #dataverse
12:35 Frank
Hi this is Frank from AUSSDA
12:35 Guest73945
Just a quick question about postgres
12:35 Guest73945
Is there a database diagram I can use to help understand the flow of data through the database?
12:43 pdurbin
Shoot, missed him. I wonder if he means something like this: http://phoenix.dataverse.org/schemaspy/latest/relationships.html
12:54 soU joined #dataverse
13:17 Frank joined #dataverse
13:17 dataverse-user joined #dataverse
13:23 icarito[m] joined #dataverse
13:24 test56465mh joined #dataverse
13:25 icarito[m]
test?
13:25 test56465mh
test2?
13:27 pdurbin_m joined #dataverse
13:27 pdurbin_m
icarito[m]: welcome!
13:27 icarito[m]
the matrix.org gateway doesn't seem to work, only one way messages
13:27 test56465mh
testing again
13:28 icarito[m]
oh I see messages both ways now
13:28 icarito[m]
hi pdurbin_m I don't see you from matrix but do from IRC
13:30 pdurbin_m
Yeah, I find Matrix confusing. I'm a simple man. I like IRC . :)
13:31 icarito[m]
yeah, it has + and - the good thing is that it integrates with other things other people use like gitter and slack, in a much simpler way than IRC
13:32 pdurbin_m
Well, Slack recently killed their IRC gateway.
13:33 dataverse-user26 joined #dataverse
13:34 dataverse-user26
Hi there, I'm trying to create an account however I'm not able to provide a password "complex" enough. I've tried different combinations of letters and numbers. Would someone be able to advise why this is the case?
13:35 dataverse-user26 left #dataverse
13:35 pdurbin_m
dataverse-user26: hi! Which installation of Dataverse are you using?
13:35 icarito[m]
anyway nice project here I may eventually put some data in dataverse for the meantime I'm loving the Dat project in general
13:39 pdurbin_m
jhand is from Dat. We love Dat.
13:40 pdurbin_m
karissa too
13:40 pdurbin_m
icarito[m]: do you think Public Lab has any research data?
13:42 icarito[m]
pdurbin_m: as the sysadmin, I think all of the photos at mapknitter.org are a good dataset to preserve - but I'm not involved with research exactly
13:42 icarito[m]
personally I wrote a Python Web IDE and am adding support for DAT to it
13:43 icarito[m]
I applied for Mozilla Open Web Fellowship with it to the Datproject, so I'm aware of jhand - twitted to him recently on it :-)
13:44 pdurbin_m
Nice, jhand visited a year or two ago and gave a great talk on Day.
13:45 pdurbin_m
icarito[m]: it would be awesome if you added support for Dataverse for your IDE too. :)
13:47 pdurbin_m
and MapKnitter looks interesting. thanks
13:47 icarito[m]
:-) yeah actually now thinking about it - it's a bit like Jupyter but without a backend - it supports transpiled Python and Markdown
13:47 pdurbin_m
dataverse-user: are you dataverse-user26? Password question?
13:48 icarito[m]
pdurbin_m: actually spectralworkbench.org also has a dataset
13:49 pdurbin_m
cool, if you know the spectralworkbench people, please feel free to reach out to them
13:49 pdurbin_m
icarito[m]: have you heard of BinderHub?
13:51 icarito[m]
I have tried some notebooks on binder but it felt slow and unreliable possibly it was my internet connection
13:51 icarito[m]
pdurbin_m: spectralworkbench.org is a publiclab.org project
13:52 icarito[m]
e.g. jywarren is the contact just like with mapknitter.org
13:52 pdurbin_m
oh! ok
14:03 pdurbin_m
icarito[m]: I opened this issue: https://github.com/jupyterhub/binder/issues/80
14:08 icarito[m]
cool. Jappy IDE is much more lightweight in that it doens't need a backend at all - but my use case is teaching coding so I haven't given much thought about processing datasets yet :-)
14:08 icarito[m]
btw Jappy IDE is available at https://educa.juegos/
14:08 icarito[m]
:-)
14:09 icarito[m]
currently it's served over Dat too but I'm adding write capabilities
14:09 icarito[m]
currently it already was using webdav experimentally for shared files / and websockets for collaborative editing
14:10 icarito[m]
a "jupyter p2p " version for researchers would be really fun
14:10 icarito[m]
eventually we'll have webasm based pythonc compliant interpreters
14:27 donsizemore joined #dataverse
14:35 pameyer joined #dataverse
15:09 pdurbin
icarito[m]: great stuff. thanks!
17:11 pdurbin
andrewSC bricas knikolla pameyer: postgres replication stuff: https://groups.google.com/d/msg/dataverse-community/TSxf4MTYYjg/RfDpMpCkAwAJ
17:25 jri joined #dataverse
17:46 pameyer
pdurbin: cool
17:49 pameyer
I remember there being a GUI option for scaling - look for a one in a circle (or a zero in a circle, before the first one has finished starting)
18:02 pdurbin
pameyer: I could swear I saw an up and down arrow at one point but I can't find it now. :/
18:04 pameyer
pdurbin: the curse of GUIs
18:04 pdurbin
I just looked again. ¯\_(ツ)_/¯
18:42 andrewSC
bomb dot com
18:51 Tyrel left #dataverse
18:54 pdurbin
andrewSC: hmm? :)
18:59 andrewSC
postgres stuff w/ kubernetes
18:59 andrewSC
looks like there's a little bit left re: your comment though
19:14 danmcp joined #dataverse
19:14 danmcp left #dataverse
19:32 pdurbin
yeah, it's just the beginning, I guess. neat stuff though
20:03 pameyer
pdurbin: do you know if any of the info load testing for openshift is anywhere in non-video format?
20:08 pdurbin
Excellent question. There's a little bit here: https://github.com/BU-NU-CLOUD-SP18/Dataverse-Scaling ... mention of ApacheBench and JMeter
20:08 pdurbin
via https://github.com/IQSS/dataverse/issues/4598#issuecomment-382489309
20:08 pdurbin
pameyer: why do you ask?
20:10 pameyer
thanks for the links - I'll check the presentations
20:10 pameyer
I was asking because of curiosity - my intuition is that glassfish was a bigger scaling bottleneck than postgres; but the pull requests started there
20:11 pameyer
since you mentioned they were doing load testing, I was hoping to see some data from them showing that my intiution was wrong; and that postgres was a bottleneck that was hit sooner than glassfish
20:12 pameyer
aka - first rule of profiling: measure it first, to figure out where the slow step is
20:15 pdurbin
I'm expecting a pull request for scaling glassfish in https://github.com/IQSS/dataverse/issues/4617 . I wouldn't sweat the order in which the pull requests came in.
20:15 pdurbin
It's always better to get your code merged first. Then the other person has to resolve merge conflicts. ;)
20:16 pdurbin
pameyer: you're such a scientist with all your measuring and data
20:17 pameyer
that too ;)
20:23 pdurbin
They're load testing Glassfish, not Postgres.
20:24 pdurbin
I bet they'd be happy to show the data. I'll ping them.
20:24 pameyer
right - but if load on glassfish doesn't translate to load on postgres, then postgres isn't the failure point
20:24 pameyer
cool - thanks
20:25 pdurbin
I think your intution is correct, that Glassfish is much more of a bottleneck that Postgres.
20:25 pdurbin
In production we've never run more than one Postgres server, I don't believe. But we had three Glassfish servers for years.
20:32 pdurbin
I emailed them and will keep you posted, pameyer
20:33 pdurbin
knikolla: I *think* they're deploying to the MOC when they run these performance tests.
20:35 pameyer
pdurbin: thanks
20:36 pameyer
any idea if MOC does openshift now? I'd thought it was openstack
20:48 pdurbin
why not both?
20:48 pdurbin
I think it's both. I'm not sure. Let's ask knikolla sometime.
20:51 pameyer
good point
21:14 jri joined #dataverse
21:14 knikolla
pameyer: pdurbin: we do openshift as well now.
21:16 pameyer
knikolla: cool, thanks
21:53 pameyer left #dataverse
22:28 pdurbin
knikolla: cool. thanks
22:28 pdurbin
Whoops, I didn't mean to say the same thing. How unoriginal of me. :)
22:32 knikolla
pdurbin: haha, for a moment i thought it was a duplicate notification.
22:33 pdurbin
heh
22:34 pdurbin
knikolla: have you tried deploying Dataverse to OpenShift? Just curious. I've only deployed to Minishift (on my laptop).
22:35 knikolla
pdurbin: i haven't
22:36 pdurbin
Ok, have you played with OpenShift much?
22:43 knikolla
pdurbin: I would really like to, however I haven't had the time.
22:44 pdurbin
Cool. I have a long list of things I don't have time to play with. :)
