Back to bug 1372589

Who When What Removed Added
Marius Cornea 2016-09-02 09:40:33 UTC CC mcornea
Vikhyat Umrao 2016-09-05 05:29:39 UTC Link ID Red Hat Knowledge Base (Solution) 1602683
Red Hat Bugzilla Rules Engine 2016-09-05 05:29:46 UTC Target Release --- 11.0 (Ocata)
Jaromir Coufal 2016-10-18 14:30:10 UTC CC jcoufal
Giulio Fidente 2016-11-16 18:23:16 UTC Blocks 1387433
CC gfidente
Assignee athomas gfidente
Giulio Fidente 2016-11-23 10:57:09 UTC CC jomurphy
Docs Contact dcadzow
QA Contact ohochman yrabl
CC twilkins
Vikhyat Umrao 2016-11-23 16:04:54 UTC Status NEW ASSIGNED
Ben England 2016-12-14 14:57:32 UTC Keywords Triaged
Blocks 1387431
Blocks 1387433
Priority unspecified medium
CC bengland
Giulio Fidente 2016-12-16 20:39:37 UTC Target Milestone --- Upstream M3
Summary [RFE] Please bump up qemu.conf the max_files to 131072 and max_processes to 65536 Please bump up qemu.conf the max_files to 131072 and max_processes to 65536
Status ASSIGNED ON_DEV
Link ID OpenStack gerrit 411983
Giulio Fidente 2016-12-16 20:51:53 UTC Link ID OpenStack gerrit 411984
Giulio Fidente 2016-12-16 20:56:57 UTC Link ID OpenStack gerrit 411987
Ben England 2016-12-16 22:29:09 UTC CC kbader, mnelson
Jeff Brown 2016-12-19 12:16:00 UTC CC jefbrown
Sean Cohen 2016-12-21 17:05:06 UTC CC scohen
Giulio Fidente 2017-01-03 16:57:38 UTC Keywords FutureFeature
Status ON_DEV POST
Ben England 2017-01-24 21:32:46 UTC Blocks 1414466
Priority medium low
Severity medium low
Priority low medium
Severity low medium
Blocks 1386905
Giulio Fidente 2017-02-02 11:25:59 UTC Status POST MODIFIED
Giulio Fidente 2017-02-06 17:05:10 UTC Component rhosp-director openstack-tripleo-heat-templates
Fixed In Version openstack-tripleo-heat-templates-6.0.0-0.20170127041112.ce54697.el7ost.1.noarch
errata-xmlrpc 2017-02-07 16:39:51 UTC Status MODIFIED ON_QA
Don Domingo 2017-02-16 04:27:30 UTC CC ddomingo
Flags needinfo?(gfidente)
Don Domingo 2017-02-16 04:47:38 UTC Docs Contact dcadzow ddomingo
Giulio Fidente 2017-02-27 21:38:15 UTC Doc Text Feature:
Allows for configuration of max_files and max_processes for the QEMU instances spawned by libvirtd via puppet hieradata. The values can be provided by the user using an enviroment file like the following:

parameter_defaults:
ExtraConfig
nova::compute::libvirt::qemu::max_files: 32768
nova::compute::libvirt::qemu::max_processes: 131072

The parameters default respectively to 32768 and 131072.

Reason:
The QEMU instances launched by libvirtd might consume a large number of file descriptors or a large number of threads depending on the number of Nova guests hosted on each compute node and of Ceph RBD images they attach to. It is necessary to be able to configure these limits in large clusters.

Result:
With the new defaults for max_files and max_processes in libvirtd, Nova computes should be able to use more than 700 OSDs, previously identified as the limit imposed by the low number of max_files (originally set to 1024, now set to 32768). Operators can customize further the value providing input as hieradata.
Flags needinfo?(gfidente)
Don Domingo 2017-02-28 01:13:47 UTC Doc Text Feature:
Allows for configuration of max_files and max_processes for the QEMU instances spawned by libvirtd via puppet hieradata. The values can be provided by the user using an enviroment file like the following:

parameter_defaults:
ExtraConfig
nova::compute::libvirt::qemu::max_files: 32768
nova::compute::libvirt::qemu::max_processes: 131072

The parameters default respectively to 32768 and 131072.

Reason:
The QEMU instances launched by libvirtd might consume a large number of file descriptors or a large number of threads depending on the number of Nova guests hosted on each compute node and of Ceph RBD images they attach to. It is necessary to be able to configure these limits in large clusters.

Result:
With the new defaults for max_files and max_processes in libvirtd, Nova computes should be able to use more than 700 OSDs, previously identified as the limit imposed by the low number of max_files (originally set to 1024, now set to 32768). Operators can customize further the value providing input as hieradata.
It is now possible to use puppet hieradata to set the max_files and max_processes for QEMU instances spawned by libvirtd. This can be done through an environment file containing the appropriate puppet classes. For example, to set the max_files and max_processes to 32768 and 131072 respectively, use:

parameter_defaults:
ExtraConfig
nova::compute::libvirt::qemu::max_files: 32768
nova::compute::libvirt::qemu::max_processes: 131072

This update also sets these values as the default, as QEMU instances launched by libvirtd might consume a large number of file descriptors or threads. This depends on Compute guest hosted on each compute node and of Ceph RBD images each instance attaches to. It is necessary to be able to configure these limits in large clusters.

With these new default values, the Compute service should be able to use more than 700 OSDs. This was previously identified as the limit imposed by the low number of max_files (originally 1024).
Don Domingo 2017-03-06 00:51:04 UTC Doc Text It is now possible to use puppet hieradata to set the max_files and max_processes for QEMU instances spawned by libvirtd. This can be done through an environment file containing the appropriate puppet classes. For example, to set the max_files and max_processes to 32768 and 131072 respectively, use:

parameter_defaults:
ExtraConfig
nova::compute::libvirt::qemu::max_files: 32768
nova::compute::libvirt::qemu::max_processes: 131072

This update also sets these values as the default, as QEMU instances launched by libvirtd might consume a large number of file descriptors or threads. This depends on Compute guest hosted on each compute node and of Ceph RBD images each instance attaches to. It is necessary to be able to configure these limits in large clusters.

With these new default values, the Compute service should be able to use more than 700 OSDs. This was previously identified as the limit imposed by the low number of max_files (originally 1024).
It is now possible to use puppet hieradata to set the max_files and max_processes for QEMU instances spawned by libvirtd. This can be done through an environment file containing the appropriate puppet classes. For example, to set the max_files and max_processes to 32768 and 131072 respectively, use:

parameter_defaults:
ExtraConfig
nova::compute::libvirt::qemu::max_files: 32768
nova::compute::libvirt::qemu::max_processes: 131072

This update also sets these values as the default, since QEMU instances launched by libvirtd might consume a large number of file descriptors or threads. This depends on Compute guest hosted on each compute node and of Ceph RBD images each instance attaches to. It is necessary to be able to configure these limits in large clusters.

With these new default values, the Compute service should be able to use more than 700 OSDs. This was previously identified as the limit imposed by the low number of max_files (originally 1024).
John Fulton 2017-03-06 12:49:54 UTC CC johfulto
Flags needinfo?(gfidente)
Giulio Fidente 2017-03-07 15:55:56 UTC Flags needinfo?(gfidente)
Giulio Fidente 2017-03-07 15:57:14 UTC Blocks 1430002
Yogev Rabl 2017-04-13 13:30:38 UTC Blocks 1414467
Status ON_QA VERIFIED
errata-xmlrpc 2017-05-17 19:32:55 UTC Status VERIFIED CLOSED
Resolution --- ERRATA
Last Closed 2017-05-17 15:32:55 UTC
Red Hat One Jira (issues.redhat.com) 2022-08-16 14:00:07 UTC Link ID Red Hat Issue Tracker OSP-4565

Back to bug 1372589