Bug 456897
| Summary: | Provide conversion to Pacemaker configuration file format | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 6 | Reporter: | Lon Hohberger <lhh> |
| Component: | pacemaker | Assignee: | Lon Hohberger <lhh> |
| Status: | CLOSED UPSTREAM | QA Contact: | Cluster QE <mspqa-list> |
| Severity: | low | Docs Contact: | |
| Priority: | high | ||
| Version: | 6.0 | CC: | cluster-maint, djansa, sghosh |
| Target Milestone: | rc | Keywords: | Reopened, TechPreview |
| Target Release: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Technology Preview | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2012-01-11 21:12:49 UTC | Type: | --- |
| 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: | 493474, 727643, 729827 | ||
| Bug Blocks: | 456895, 756082 | ||
|
Description
Lon Hohberger
2008-07-28 14:42:46 UTC
I have this working for limited sets of resources and STONITH configurations. The following is known to work:
* single-instance STONITH devices for a host
* All parameters must be in the <fencedevice> tag
* multiple-instance STONITH devices with or without hostlist support
* caveat: you must only have 'port="X"' in the <device> tag; no other
options from the <device> tag are handled; they must be in the
<fencedevice> tag
* IP address resources
* script resources
I tested:
1) Services running under rgmanager; both cluster nodes online
2) Run ccs2cib
[root@snap ccs2cib]# ./ccs2cib -d -o cib.xml -R
debug: adding . to /usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin:/root/bin
debug: using ./cluconf2cib.xsl
debug: using /usr/share/java/saxon.jar
* Converting configuration
* Calling crm_verify to validate the configuration.
* Disabling rgmanager in /etc/cluster/cluster.conf
* Calling cman_tool to update cluster.conf
* Proposed cib stored in cib.xml
[root@snap ccs2cib]#
3) Wait for rgmanager to finish exiting on both hosts
4) Start pacemaker on both hosts
5) cibadmin --replace --xml-file cib.xml
NOTE: I had to do this twice; I am not sure why.
Check crm_mon:
[root@snap ccs2cib]# crm_mon -1
============
Last updated: Wed Aug 10 15:00:17 2011
Stack: cman
Current DC: crackle - partition with quorum
Version: 1.1.5-5.el6-01e86afaaa6d4a8c4836f68df80ababd6ca3902f
3 Nodes configured, unknown expected votes
4 Resources configured.
============
Online: [ snap crackle ]
OFFLINE: [ /dev/block/8:1 ]
Resource Group: service_1
ip_192_168_122_201 (ocf::redhat:ip.sh): Started crackle
Resource Group: service_dang
script_uhoh (ocf::redhat:script.sh): Started snap
st_virt (stonith:fence_xvm): Started crackle
6) Kill 'snap' with 'reboot -fn'
7) Check crm_mon on 'crackle':
============
Last updated: Wed Aug 10 15:01:16 2011
Stack: cman
Current DC: crackle - partition with quorum
Version: 1.1.5-5.el6-01e86afaaa6d4a8c4836f68df80ababd6ca3902f
3 Nodes configured, unknown expected votes
4 Resources configured.
============
Online: [ crackle ]
OFFLINE: [ snap /dev/block/8:1 ]
Resource Group: service_1
ip_192_168_122_201 (ocf::redhat:ip.sh): Started crackle
Resource Group: service_dang
script_uhoh (ocf::redhat:script.sh): Started crackle
st_virt (stonith:fence_xvm): Started crackle
Failed actions:
vm_foo_monitor_0 (node=crackle, call=4, rc=5, status=complete): not installe
d
Oops -- the vm_foo_monitor_0 action was present in the crm_mon on snap, too, but it's a dummy resource and can be ignored. I have tagged ccs2cib v0.5; this can be pulled in to other projects as needed. It still needs a bit more polish and testing, which will be performed in the mainline / master branch for now. https://github.com/lhh/ccs2cib/tree/v0.5 |