On Fri, Dec 24, 2010 at 08:59:29AM +0200, Alexander Bodnarashik wrote:
>
> On Dec 23, 2010, at 18:53, Dejan Muhamedagic wrote:
>
> > On Fri, Dec 10, 2010 at 02:08:34PM +0200, Alexander Bodnarashik wrote:
> >> I've moved a little bit further with running linux ha suite on freebsd,
> >> few more problems though:
> >>
> >> 1. crm node standby/online is not working
> >> issuing following command has no effect and node remains online/standby
> >>> [r...@alice ~]# crm node standby alice.local reboot
> >>> scope=nodes name=standby value=off
> >>> [r...@alice ~]#
> >> debug log says:
> >>> Dec 10 13:51:09 alice cib: [50795]: debug: cib_process_xpath: Processing
> >>> cib_query op for
> >>> //cib/configuration/nodes//no...@id='alice.local']//nvpa...@name='standby']
> >>> (/cib/configuration/nodes/node[3]/instance_attributes/nvpair)
> >>> Dec 10 13:51:09 alice cib: [50795]: debug: cib_process_xpath: Processing
> >>> cib_query op for
> >>> //cib/configuration/nodes//no...@id='alice.local']//nvpa...@name='standby']
> >>> (/cib/configuration/nodes/node[3]/instance_attributes/nvpair)
> >> in case if attribute does not exist output is following:
> >>> [r...@alice ~]# crm node standby alice reboot
> >>> Error performing operation: The object/attribute does not exist
> >>> scope=status name=standby value=off
> >> and logs:
> >>> Dec 10 14:07:36 alice cib: [50795]: debug: cib_process_xpath: cib_query:
> >>> //cib/configuration/nodes//no...@id='alice.local']//nvpa...@name='standby']
> >>> does not exist
> >>> Dec 10 14:07:36 alice cib: [50795]: debug: cib_process_xpath: cib_query:
> >>> //cib/status//node_sta...@id='alice.local']//nvpa...@name='standby'] does
> >>> not exist
> >>
> >> on the other hand following puts node in standby mode:
> >>> [r...@alice ~]# crm node attribute alice.local set standby on
> >>> [r...@alice ~]#
> >
> > Best to open a bugzilla for this. It's funny because newer crm
> > shell version in 1.1 uses crm_attribute for standby/online. The
> > stable 1.0 series however still relies on crm_standby.
> I guess i will, wanna try pacemaker 1.1 before.
crm_standby is available in 1.1 too. Andrew modified the shell
to use crm_attribute in its stead. The log message says:
Use crm_attribute directly instead of the
deprecated crm_standby and crm_failcount wrappers
> >> 2. hb_report basically does not work as expected:
> >>> [r...@alice ~]# hb_report -f 1pm /root/hb.log
> >>> alice.local: INFO: logd used but logd.cf not found: using defaults
> >>> alice.local: WARN: this is not a node and you didn't specify a list of
> >>> nodes using -n
> >>> alice.local: WARN: could not find on alice.local
> >
> > hb_report obviously cannot figure out that this node is a member
> > of the cluster. Was pacemaker running at that point in time? Did
> > you try to specify the nodes with -n?
> Yes, pacemaker was running on the node.
> I tried to specify nodes with -n:
> [r...@alice ~]# hb_report -f 1pm -n alice.local /root/hb.log
> alice.local: INFO: logd used but logd.cf not found: using defaults
> alice.local: WARN: could not find on alice.local
> usage: mktemp [-d] [-q] [-t prefix] [-u] template ...
> mktemp [-d] [-q] [-u] -t prefix
> alice.local: WARN: sorry, can't create temporary file for find_files
> usage: mktemp [-d] [-q] [-t prefix] [-u] template ...
> mktemp [-d] [-q] [-u] -t prefix
> alice.local: WARN: sorry, can't create temporary file for find_files
> usage: mktemp [-d] [-q] [-t prefix] [-u] template ...
> mktemp [-d] [-q] [-u] -t prefix
> alice.local: ERROR: cannot create temporary files
Oops. mktemp has different usage in *BSD. We need to specify the
template.
> [r...@alice ~]# crm_mon -1
> ============
> Last updated: Fri Dec 24 08:52:38 2010
> Stack: openais
> Current DC: bob.local - partition with quorum
> Version: 1.0.10-b2e39d318fda501e2fcf223c2d039b721f3679a9
> 3 Nodes configured, 3 expected votes
> 1 Resources configured.
> ============
>
> Online: [ mike.local bob.local alice.local ]
>
> Resource Group: gr
> dummy1 (ocf::heartbeat:Dummy): Started mike.local
> dummy2 (ocf::heartbeat:Dummy): Started mike.local
> dummy3 (ocf::heartbeat:Dummy): Started mike.local
> dummy4 (ocf::heartbeat:Dummy): Started mike.local
> dummy5 (ocf::heartbeat:Dummy): Started mike.local
> dummy6 (ocf::heartbeat:Dummy): Started mike.local
> dummy7 (ocf::heartbeat:Dummy): Started mike.local
> dummy8 (ocf::heartbeat:Dummy): Started mike.local
> dummy9 (ocf::heartbeat:Dummy): Started mike.local
> dummy10 (ocf::heartbeat:Dummy): Started mike.local
> dummy11 (ocf::heartbeat:Dummy): Started mike.local
> dummy12 (ocf::heartbeat:Dummy): Started mike.local
> dummy13 (ocf::heartbeat:Dummy): Started mike.local
> dummy14 (ocf::heartbeat:Dummy): Started mike.local
> dummy15 (ocf::heartbeat:Dummy): Started mike.local
> [r...@alice ~]#
>
> At this point there is nothing suspicious in logs:
> Dec 24 08:52:18 alice cib: [886]: WARN: adjust rcvbuf size to 1048576 failed:
> No buffer space available
> Dec 24 08:52:18 alice cib: [886]: WARN: adjust sndbuf size to 1048576 failed:
> No buffer space available
> Dec 24 08:52:18 alice cib: [886]: WARN: adjust rcvbuf size to 1048576 failed:
> No buffer space available
> Dec 24 08:52:18 alice cib: [886]: WARN: adjust sndbuf size to 1048576 failed:
> No buffer space available
>
> >
> >>> The report is saved in /root/hb.log.tar.bz2
> >>> [r...@alice ~]# tar tjf /root/hb.log.tar.bz2
> >>> hb.log/
> >>> hb.log/analysis.txt
> >>> hb.log/description.txt
> >>> [r...@alice ~]#
> >>
> >>
> >> My configuration:
> >>> [r...@alice ~]# uname -a
> >>> FreeBSD alice.local 8.1-RELEASE FreeBSD 8.1-RELEASE #0: Mon Jul 19
> >>> 02:55:53 UTC 2010
> >>> [email protected]:/usr/obj/usr/src/sys/GENERIC i386
> >>
> >>> [r...@alice ~]# cibadmin -$
> >>> cibadmin 1.0.10 for OpenAIS (Build:
> >>> b0266dd5ffa9c51377c68b1f29d6bc84367f51dd)
> >>>
> >>> Written by Andrew Beekhof
> >>
> >>> [r...@alice ~]# corosync -v
> >>> Corosync Cluster Engine, version '1.2.8' SVN revision '3035'
> >>> Copyright (c) 2006-2009 Red Hat, Inc.
> >>
> >>> [r...@alice ~]# crm configure show
> >>> node alice.local \
> >>> attributes standby="on"
> >>> node bob.local \
> >>> attributes standby="off"
> >>> node mike.local
> >>> property $id="cib-bootstrap-options" \
> >>> dc-version="1.0.10-b0266dd5ffa9c51377c68b1f29d6bc84367f51dd" \
> >>> cluster-infrastructure="openais" \
> >>> expected-quorum-votes="3" \
> >>> no-quorum-policy="ignore" \
> >>> stonith-enabled="true" \
> >>> last-lrm-refresh="1291296991" \
> >>> default-resource-stickiness="100" \
> >>> stonith-action="reboot"
> >>> rsc_defaults $id="rsc-options" \
> >>> allow-migrate="true"
> >>> op_defaults $id="op-options" \
> >>> on-fail="restart"
> >>
> >>> [r...@alice ~]# crm_mon -1
> >>> ============
> >>> Last updated: Fri Dec 10 14:02:28 2010
> >>> Stack: openais
> >>> Current DC: alice.local - partition WITHOUT quorum
> >>> Version: 1.0.10-b0266dd5ffa9c51377c68b1f29d6bc84367f51dd
> >>> 3 Nodes configured, 3 expected votes
> >>> 0 Resources configured.
> >>> ============
> >>>
> >>> Node alice.local: standby
> >>> OFFLINE: [ mike.local bob.local ]
> >>
> >> Please let me know if i should provide more info.
> >
> > Are those nodes really offline? Did you manage to form a
> > cluster?
> >
> Yes, they were really offline. I've switched them off to minimize test case.
> I've managed to form a cluster, nodes see each other, failover/fencing
> (meatware) seems to work, but i have not tested cluster much.
Great. If you feel like testing, there's CTS.
Thanks,
Dejan
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
