Bug 1329302

Summary:	Default min_command_interval is too low, results in nodes losing connectivity during introspection
Product:	Red Hat OpenStack	Reporter:	Ronelle Landy <rlandy>
Component:	openstack-ironic	Assignee:	Lucas Alvares Gomes <lmartins>
Status:	CLOSED WORKSFORME	QA Contact:	Raviv Bar-Tal <rbartal>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	8.0 (Liberty)	CC:	dtantsur, dyocum, mburns, rhel-osp-director-maint, rlandy, srevivo, whayutin
Target Milestone:	---	Keywords:	Automation, Documentation, Reopened, ZStream
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-11-10 15:48:44 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Comment 2 Dan Yocum 2016-04-21 15:05:06 UTC

The proper parameter to set is in /etc/ironic/ironic.conf:

[ipmi]
min_command_interval=10

Then restart the ironic services 'systemctl restart openstack-ironic*'

The puppet modules for ironic should put this parameter into the config file (with the default value) so the operator can modify it as they see fit.

Furthermore, I suggest that ALL parameters should be included in the ironic.conf file for operator reference.  The comments would be nice, too.

Comment 3 Lucas Alvares Gomes 2016-05-19 10:10:27 UTC

Hi Ronelle, Dan,

This varies depending on hardware and operators should adjust it according to what they have, unfortunately there's no perfect value that will work for everybody.

By default Ironic sets 5 seconds between each IPMI command (per BMC), which is already a lot in computing terms.

Comment 6 Dmitry Tantsur 2016-11-10 11:53:43 UTC

Hi! Is this still a problem? I don't think that bumping that ironic default is a way to go, but we can probably bump it via puppet for tripleo specifically.

Comment 7 Ronelle Landy 2016-11-10 15:36:09 UTC

Seems to be working now as far as the CI machines are concerned.

Comment 8 Dmitry Tantsur 2016-11-10 15:48:44 UTC

Thanks! I'm closing it for now. If somebody sees similar issues, we can reopen it and change the option via puppet.