Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
Jason Frey
@Fryguy
I think this was fixed, right? but probably not released yet?
Adam Grare
@agrare
correct it is fixed in the morphy branch already, just pending morphy-1 release
wallerz
@wallerz
@agrare @Fryguy 修复了我去哪里进行下载安装?
cymile
@cymile

Hi! I have a deployment of manageIQ ivanchuk-7 with postgresql ( pgpool + posgresql/repmgr) in a kubernetes cluster..all pods in the same node

NAME                                             READY   STATUS    RESTARTS   AGE   IP             NODE    NOMINATED NODE   READINESS GATES
manageiq-74fccdb898-crhn4                        1/1     Running   0          9h    10.233.92.32   node3   <none>           <none>
postgres-postgresql-ha-pgpool-84b46b758f-bxfx8   1/1     Running   0          10h   10.233.92.30   node3   <none>           <none>
postgres-postgresql-ha-postgresql-0              1/1     Running   0          10h   10.233.92.29   node3   <none>           <none>

and after some time I got 503 error using the UI. And the logs I got are:
In ManageIQ

[----] I, [2021-11-11T13:48:53.607610 #220:9ee5f8]  INFO -- : Q-task_id([job_dispatcher]) MIQ(InfraConversionThrottler.running_conversion_jobs) Running InfraConversionJob: 0
[----] I, [2021-11-11T13:48:53.618966 #220:9ee5f8]  INFO -- : Q-task_id([job_dispatcher]) MIQ(JobProxyDispatcher#dispatch) Complete - Timings: {:v2v_dispatching=>0.0039370059967041016, :v2v_limits=>0.0039043426513671875, :pending_container_jobs=>0.004796743392944336, :container_jobs_to_dispatch_count=>0, :container_dispatching=>0.004833221435546875, :pending_vm_jobs=>0.0010466575622558594, :vm_jobs_to_dispatch_count=>0, :total_time=>0.016845226287841797}
[----] I, [2021-11-11T13:48:53.619457 #220:9ee5f8]  INFO -- : Q-task_id([job_dispatcher]) MIQ(MiqQueue#delivered) Message id: [4330], State: [ok], Delivered in [0.017842446] seconds
[----] I, [2021-11-11T13:48:55.407403 #8205:9ee5f8]  INFO -- : MIQ(MiqReportingWorker::Runner#message_sync_config) MIQ(MiqReportingWorker::Runner) Synchronizing configuration...
[----] E, [2021-11-11T13:48:55.411007 #8205:9ee5f8] ERROR -- : MIQ(MiqReportingWorker::Runner) ID [115] PID [8205] GUID [80b29f9a-a86c-417c-94be-946217eae4b4] Error in update_worker_record_at_exit: PG::UnableToSend: no connection to the server
: SELECT  "miq_workers".* FROM "miq_workers" WHERE "miq_workers"."type" IN ('MiqReportingWorker') AND "miq_workers"."id" = $1 LIMIT $2
[----] E, [2021-11-11T13:48:55.412401 #8205:9ee5f8] ERROR -- : MIQ(MiqReportingWorker::Runner) ID [115] PID [8205] GUID [80b29f9a-a86c-417c-94be-946217eae4b4] Error heartbeating because ActiveRecord::StatementInvalid: PG::ConnectionBad: PQconsumeInput() server closed the connection unexpectedly
    This probably means the server terminated abnormally
    before or while processing the request.

In PGPOOL

2021-11-11 13:49:17: pid 17467: ERROR: unable to read data 2021-11-11 13:49:17: pid 17467: DETAIL: child connection forced to terminate due to client_idle_limit:3600 is reached 2021-11-11 13:49:18: pid 1980: ERROR: unable to read data 2021-11-11 13:49:18: pid 1980: DETAIL: child connection forced to terminate due to client_idle_limit:3600 is reached
is there a way to modify how manageiq handle database connections? It seems it is still using a dead connection o some help around this?

Jason Frey
@Fryguy
@wallerz Once we release beta1, we will announce it and it will be on the downloads page - hoping today
Daniel Berger
@djberg96
@cymile Found this, not sure if it's something you can configure within ManageIQ though: https://b-peng.blogspot.com/2020/07/connection-pooling-in-pgpool.html
Scottwake3
@Scottwake3
Hello, is there a method for only allowing subscriptions to be deleted once they are empty?
Alejandro Macedo
@AMacedoP
Hi, while using the embedded ansible I run into playbooks that fail with timeouts because they take a long time to complete. I tried searching for an option to increase it, but it seems that it uses the generic job timeout option, however increasing it didn't change anything. Is there an option to only increase the embedded ansible timeout?
Peter McGowan
@pemcg
@AMacedoP there’s a TTL (time to live) setting on the playbook method and service definition page, have you tried increasing that?
Alejandro Macedo
@AMacedoP
@pemcg thanks so much, didn't there was a TTL there
Aleksei Pilnik
@apilnik
Hello to everyone!
please tell me is it possible that the placement of the virtual machine depends on its name, if so, how to do it?
Peter McGowan
@pemcg
@Alexis3211 placement is designed to be user customisable, and yes you can do exactly what you want. Checkout the methods here for examples: /Infrastructure/VM/Provisioning/Placement/
Aleksei Pilnik
@apilnik
@pemcg Can you help me with the code please? Because I'm not good enough at it ((
Peter McGowan
@pemcg
@Alexis3211 Have a look through the forum, there are several examples of placement code: https://talk.manageiq.org/search?q=placement
Aleksei Pilnik
@apilnik
@pemcg thank you the best:)
Javier Richard Quinto Ancieta
@richardqa
Hi everyone, I have a question, ManageIQ support Mysql or MariaDB as backend?
Aleksei Pilnik
@apilnik
@pemcg Hi! I try to integrate Ansible Tower to ManageIQ. I've already connected provider and make catalog bundle, and it even works!!! Can we somehow automatically add new host to tower inventory?
vasilatos v
@vasgitlab_gitlab
Hello everyone
I have installed gitlab gem for my own plugin development
but I can not load it in my controller
Are there any solutions?
```[----] F, [2021-11-28T02:11:31.957501 #740673:e100] FATAL -- : Error caught: [NameError] uninitialized constant GitlabController::Gitlab
```
vasilatos v
@vasgitlab_gitlab
[----] F, [2021-11-28T02:11:31.957501 #740673:e100] FATAL -- : Error caught: [NameError] uninitialized constant GitlabController::Gitlab
Hello @Fryguy
Jason Frey
@Fryguy
@richardqa No, we use some Postgresql specific features, and can't use mysql anymore
@vasgitlab_gitlab How did you include it? It needs to be in your .gemspec
Adam Grare
@agrare
@vasgitlab_gitlab looks like you need to require it
vasilatos v
@vasgitlab_gitlab
okay thank you for your information
cymile
@cymile
Hi there! is there a current integration with Ansible Tower v.3.8 ? I checked Ansible Tower documentation for this version and saw there is not Cloudform credential type support anymore
vasilatos v
@vasgitlab_gitlab
I faced error
not found such file -- util/vmdb-logger
during bin/update
Adam Grare
@agrare
@vasgitlab_gitlab make sure your core repository is up to date (e.g. git pull upstream master)
Jay Carman
@jaywcarman

I pulled the latest core master this morning (80714fcf18), updated everything else, and now am unable to authenticate to the web interface. Seeing the following error:

[----] F, [2021-12-06T07:58:11.743296 #679013:e199408] FATAL -- development: Error caught: [TypeError] can't cast Rack::Session::SessionId

Trying to figure out if this is something local to my env or if I pulled in a bug when I updated.

Adam Grare
@agrare
@jaywcarman that's not just you I'm also seeing that
Jay Carman
@jaywcarman

Pushed full logs up to the tmp branch:
https://raw.githubusercontent.com/jaywcarman/manageiq/log_tmp/log.development.20211206.log

@agrare nice to know it's not something I goofed :)

Adam Grare
@agrare
there was a new version of the dalli gem released on friday (3.1.0)
you can lock down to ~> 3.0.0 until we fix it
Gemfile
-gem "dalli",                            "~>3.0",             :require => false
+gem "dalli",                            "~>3.0.0",             :require => false
Jay Carman
@jaywcarman
Perfect, thanks!
Jason Frey
@Fryguy
dalli bumped - we noticed on friday and have to lockdown - cc @bdunne / @jrafanie
Brandon Dunne
@bdunne
oh no
Jason Frey
@Fryguy
I think they changed the method we patched
Adam Grare
@agrare

https://github.com/petergoldstein/dalli/blob/main/History.md

BREAKING CHANGE .... minor version bump :)

Jason Frey
@Fryguy
ugh - glad they don't follow semver
Adam Grare
@agrare
looking back they only did a major version bump for serious overhauls, they do minor bumps for breaking changes
might want to consider ~> to the patch version
Jason Frey
@Fryguy
yeah was just thinking that
maybe a comment too in the Gemfile that they don't follow semver properly