Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1714816

Summary: [OSP13] sshd running inside nova_migration_target overrides /var/run/sshd.pid on host with pid 1
Product: Red Hat OpenStack Reporter: Takashi Kajinami <tkajinam>
Component: openstack-tripleo-heat-templatesAssignee: Piotr Kopec <pkopec>
Status: CLOSED ERRATA QA Contact: Martin Schuppert <mschuppe>
Severity: high Docs Contact:
Priority: medium    
Version: 13.0 (Queens)CC: amodi, knoha, mbooth, mburns, mschuppe
Target Milestone: z8Keywords: Triaged, ZStream
Target Release: 13.0 (Queens)   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: openstack-tripleo-heat-templates-8.3.1-57.el7ost Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
: 1724674 (view as bug list) Environment:
Last Closed: 2019-09-03 16:55:32 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:
Embargoed:
Bug Depends On:    
Bug Blocks: 1724674    

Description Takashi Kajinami 2019-05-28 23:45:43 UTC 
Description of problem:

In RHOSOP13, we have 2 sshd instances running on compute nodes.
 1. sshd running on host, to use conventional operation over remote login
 2. sshd running inside nova_migration_target, used for migration in nova

These two instances generally use separated resources, like ports, or conf files,
but they are sharing pid, /var/run/sshd.pid on the host.

This causes that we have "1" in /var/run/sshd.pid, as pid file is overwritten
by the sshd running inside nova_migration_target container with pid 1 inside the container.

Note that currently we do not see any specific problem caused by this,
except for the below error log shown when we restart sshd running on host.
~~~
[heat-admin@compute-0 ~]$ sudo systemctl restart sshd
[heat-admin@compute-0 ~]$ sudo systemctl status sshd
● sshd.service - OpenSSH server daemon
...
May 28 23:31:45 compute-0 sshd[422824]: error: Couldn't create pid file "/var/run/sshd.pid": Permission denied
~~~

Version-Release number of selected component (if applicable):
RHOSP13z6

How reproducible:
Always

Steps to Reproduce:
1. Restart nova_migration_target container
2. Check content in /var/run/sshd.pid

Actual results:
We have pid 1 in the pid file

Expected results:
We have pid for sshd running at host level in the pid file

Additional info:
Comment 1 Keigo Noha 2019-05-29 07:29:11 UTC 
The cause of this issue is ./docker/services/nova-migration-target.yaml contains following volume mount.
~~~
    145             volumes:
    146               list_concat:
    152                   - /run:/run
~~~
In host side, /var/run is a symbolic link to /run and the directory is used by the host side sshd also.
Comment 4 Martin Schuppert 2019-06-26 12:52:10 UTC 
This introduced an issue with live migration as the nova-migration-wrapper inside the container needs access to the libvirt socket [1].

We track the fix for this in https://bugzilla.redhat.com/show_bug.cgi?id=1724131

[1] https://github.com/rdo-packages/nova-distgit/blob/rpm-master/nova-migration-wrapper#L31
Comment 16 errata-xmlrpc 2019-09-03 16:55:32 UTC 
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2019:2624