[ClusterLabs] Antw: Re: big trouble with a DRBD resource
Lentes, Bernd
bernd.lentes at helmholtz-muenchen.de
Mon Aug 7 11:19:08 EDT 2017
----- On Aug 7, 2017, at 3:43 PM, Ulrich Windl Ulrich.Windl at rz.uni-regensburg.de wrote:
>>
>>>
>>> The "ERROR" message is coming from the DRBD resource agent itself, not
>>> pacemaker. Between that message and the two separate monitor operations,
>>> it looks like the agent will only run as a master/slave clone.
>>
>> Btw:
>> Does the crm/lrm call the RA in the same way init does with scripts in
>> /etc/init.d ?
>
> Hi Bernd!
>
> If you ever tried to run an RA by hand, you wouldn't have asked, I guess.
> The RA needs ist parameters from environment, while typical LSB scripts take
> their configuration from files (/etc/sysconfig). I hope you are not talking
> about using LSB scripts as RAs.
I'm not talking about LSB scripts as RA's, although that's possible:
crm(live)# ra info lsb:
Display all 126 possibilities? (y or n)
lsb:SuSEfirewall2_init lsb:boot.debugfs lsb:boot.multipath lsb:elx-lpfc-vector-map lsb:irq_balancer lsb:networker lsb:purge-kernels lsb:snmpd
lsb:SuSEfirewall2_setup lsb:boot.device-mapper lsb:boot.open-iscsi lsb:fbset lsb:joystick lsb:nfs lsb:radvd lsb:snmptrapd
lsb:alsasound lsb:boot.dmraid lsb:boot.proc lsb:gpm lsb:kbd lsb:nfsserver lsb:random lsb:spamd
lsb:arpd lsb:boot.efivars lsb:boot.rootfsck lsb:haldaemon lsb:kexec lsb:nscd lsb:raw lsb:splash
lsb:atd lsb:boot.fuse lsb:boot.swap lsb:halt lsb:ksysguardd lsb:ntp lsb:rc lsb:splash_early
lsb:atop lsb:boot.ipconfig lsb:boot.sysctl lsb:halt.local lsb:ldirectord lsb:nxsensor lsb:reboot lsb:sshd
lsb:auditd lsb:boot.kdump lsb:boot.sysstat lsb:haveged lsb:libvirt-guests lsb:nxserver lsb:rpasswdd lsb:syslog
lsb:autoyast lsb:boot.klog lsb:boot.udev lsb:hawk lsb:libvirtd lsb:o2cb lsb:rpcbind lsb:tg3sd
lsb:bluez-coldplug lsb:boot.ldconfig lsb:boot.udev_retry lsb:hp-ams lsb:lighttpd lsb:ocfs2 lsb:rpmconfigcheck lsb:uuidd
lsb:boot lsb:boot.loadmodules lsb:cron lsb:hp-asrd lsb:logd lsb:open-iscsi lsb:rsyncd lsb:virtlockd
lsb:boot.cgroup lsb:boot.local lsb:ctdb lsb:hp-health lsb:lvm_wait_merge_snapshot lsb:openais lsb:setserial lsb:winbind
lsb:boot.cleanup lsb:boot.localfs lsb:dbus lsb:hp-snmp-agents lsb:mdadmd lsb:openhpid lsb:sfcb lsb:xdm
lsb:boot.clock lsb:boot.localnet lsb:dnsmasq lsb:hpsmhd lsb:multipathd lsb:pcscd lsb:single lsb:xfs
lsb:boot.compliance lsb:boot.lvm lsb:drbd lsb:ipmi lsb:mysql lsb:postfix lsb:skeleton lsb:xinetd
lsb:boot.crypto lsb:boot.lvm_monitor lsb:earlysyslog lsb:ipmievd lsb:network lsb:powerd lsb:skeleton.compat
lsb:boot.crypto-early lsb:boot.md lsb:ebtables lsb:ipvsadm lsb:network-remotefs lsb:powerfail lsb:smbfs
Calling a RA by hand is recommended for debugging:
https://wiki.clusterlabs.org/wiki/Debugging_Resource_Failures
>> E.g if a resource has a monitor interval from 120, the RA is called every 2
>> min from the cluster with the option monitor ?
>
> I hope you are not expecting init (etc/init.d) to do periodic status checks
> (actually init can check the status of processes it stared directly, but I'm
> sure you know that).
No. I didn't mention this.
Bernd
Helmholtz Zentrum Muenchen
Deutsches Forschungszentrum fuer Gesundheit und Umwelt (GmbH)
Ingolstaedter Landstr. 1
85764 Neuherberg
www.helmholtz-muenchen.de
Aufsichtsratsvorsitzende: MinDir'in Baerbel Brumme-Bothe
Geschaeftsfuehrer: Prof. Dr. Guenther Wess, Heinrich Bassler, Dr. Alfons Enhsen
Registergericht: Amtsgericht Muenchen HRB 6466
USt-IdNr: DE 129521671
More information about the Users
mailing list