Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Scott Williams
@vwbusguy
Looks like that fixed it!
Many thanks @simaishi , @NickLaMuro , @Fryguy !
It seems that I needed to have qpid-proton-c{,devel} installed, and specifically < 0.30.0
So something changed in ivanchuk-7 in the openstack that causes qpid-proton-c to now be required where it used to be optional. I'm not actually using the openstack provider for anything, which might be why I never bothered to install it before.
About it being required now, should I file an issue for it in manageiq-providers-openstack?
Satoe Imaishi
@simaishi
If it's trying to load qpid_proton when it's not needed, it's an issue in openstack provider. if you're using feature that requires it (openstack event monitor?), but didn't install it because it's not documented, then it's a documentation bug, I think... @agrare is that right?
Adam Grare
@agrare

That line itself seems the same as ivanchuk-6

The difference is the STF event monitor was added in ivanchuk-7

Scott Williams
@vwbusguy
So it seems like a documentation issue then, as qpid-proton-c <0.30.0 is now required
I wasn't using any openstack features, but evm failed to start without it regardless
Adam Grare
@agrare
It is something we could handle but we never expected to be running the full-up server without qpid-proton installed
being able to run without it is primarily for development purposes
Scott Williams
@vwbusguy
So it sounds like it should probably just be required then?
Satoe Imaishi
@simaishi
< 0.30.0 is only for ivanchuk, fyi
different branch has different version requirement
Adam Grare
@agrare
we could rescue qpid_proton (LoadError) here but we would also have to make sure to not present STF as an option in the dropdown for event catcher types on the UI/API
Scott Williams
@vwbusguy
We're planning on moving to containers for jansa anyway, so I imagine those deps will be handled at the image level at that point
This is one of those cases where the demo environment turned into the prod environment because a bunch of unexpected things happened that drew our resources elsewhere...
Adam Grare
@agrare
correct
Scott Williams
@vwbusguy
But the upside is, I get to learn more about how MIQ works this way :-)
Adam Grare
@agrare
:laughing: I like the optimism!
Scott Williams
@vwbusguy
We were evaluating buying vRealize stuff, so I spun up ManageIQ just to get a sense of what a competing platform was like and generally, most ended up liking ManageIQ much more. Then company financial problems and COVID happened, so we ended up not being able to afford vRealize anyway, but we were already leaning toward MIQ at that point.
Jason Frey
@Fryguy

and generally, most ended up liking ManageIQ much more

💕💖

that made my week :smile:
Nick LaMuro
@NickLaMuro
Scott Williams
@vwbusguy
@Fryguy :-D It really is a great platform.
pyon-yx
@pyon-yx
Is there anybody who can teach me how to integrate the openstack object storage to manageiq?
Adam Grare
@agrare
The cinder manager should be added automatically, what are you having problems with?
pyon-yx
@pyon-yx
@agrare Thanks for reply.
Yes, you're right.
cinder manager is working well.
I want to know how swift manager would be working .
TobyJetson
@TobyJetson
Hello, is anyone available on here? I am trying to work out if it's possible to pass the username of the currently-logged-in ManageIQ user down to an embedded Ansible playbook
Peter McGowan
@pemcg
@TobyJetson An href_ slug to the user’s ID is already passed in as {{ manageiq.user }}, so you could either make a quick API lookup on that slug to get the username, or alternatively just pass in ${/#user.userid} as a string input parameter if it’s a playbook method
Adam Grare
@agrare
Swift manager should be automatically set up as well
Keenan Brock
@kbrock
If people are curious about cross bot testing, the test by @NickLaMuro are some beautiful reading https://github.com/ManageIQ/miq_bot/blob/3bf7c223ae241db26f69bc0c8a93ece7ca6a2b56/spec/lib/github_service/commands/cross_repo_test_spec.rb
Keenan Brock
@kbrock
some said there was an issue with spec/lib/api/api_config_spec.rb:39 - @agrare was that you?
think this has something to do with the caching changes I introduced around miq features and tenancy
Nick LaMuro
@NickLaMuro
@kbrock see the API channel
I just posted where the root problem was (and it isn't that)
Keenan Brock
@kbrock
sweet
p-v-a
@p-v-a

Anyone seen problem with embedded ansible in kubernetes deployment? I'm getting and error when I try to run playbook. it complains about permissions on /root/.ansible. If I exec into container I can see that indeed /root is missing write permissions.

[root@1-generic-8d88f4d7-v9vnz /]# ls -lad /root/
dr-xr-x---. 1 root root 4096 Nov 17 08:47 /root/

thus ansible can't create .ansible directory inside root. there are few possible solutions for that issue:

thing is can't see where it's coming from though, there is no reports of this issue on github issues, so I'm not sure if this is something about my particular deployment or it's a genuine issue. Any thoughts?

p-v-a
@p-v-a
After some digging I ended up with the following:
  • it's known issue in centos,rhel - CentOS/sig-cloud-instance-images#67
  • my containers were running with requiredDropCapabilities: ["ALL"], depriving container of CAP_FOWNER that would allow bypass permission checks on operations that normally require the filesystem UID of the process to match the UID of the file, thus allowing access even owner permissions is wrong.
Jason Frey
@Fryguy
interesting...thanks for figuring that out @p-v-a
SeungWookJang
@SeungWookJang
I got an error when I add a new cloud and automate provider in MIQ developer setup.
"timed out stalled task"
Is there anyone who can help me?
Adam Grare
@agrare
@SeungWookJang that means that nothing processed the async queue item that verifies the credentials within the timeout period (1 minute I think)
if this was a development setup are you running simulate_queue_worker ?
swjang
@swjang
@agrare Yes, it’s a developer setup. What is the simulate_queue_worker?
Adam Grare
@agrare
@swjang it is a lightweight way of running MIQ for development without starting the full server, basically it just dequeues all queue items typically handled but a bunch of various worker types
rails s in one terminal and simulate_queue_worker in rails c in another terminal is a dev technique, not everything will work but a lot will