I thought i'd fixed it, but now I know I have. Latest CVS will work.
On 1/26/06, Andrew Beekhof <[EMAIL PROTECTED]> wrote: > > On Jan 26, 2006, at 9:32 AM, Francis Montagnac wrote: > > > > >>> Well he IS sending in the full parent nodes. > > > >> ahh, i didnt notice the grep wasnt in the second command > > > > Sorry. I should have been more precise. > > > >>> He's basically feeding back an entire CIB with just that change, > > > > Yes. > > > >>> and I think that includes feeding back various state data etc. > >>> Mayb'e > >>> that's screwing up stuff? > > > > May be not. See below. > > > > What I would like to do is to have the static part of the CIB, (ie: > > crm_config, resources and constraints) in one single file and be > > able to > > update the CIB from this file. > > > > Am I wrong to want that? > > The short answer to that is yes i'm afraid. > > > > > Reading more the cibadmin usage I tried to: > > > > - add an ID to the cib tag and to every tag till > > transition_idle_timeout > > You cant do that. ID fields are immutable, different values make it > a different object. > So if you want to "change" it, you have to delete the old object and > add a new one with the new ID. > > > - call cibadmin --id <this ID> -U > > This is what the attachment corresponds to? > > > > > This time no errors at all, but the update (same as before: > > changing the > > transition_idle_timeout value as a test) wasn't done. > > > > Here are the commands I tried and the full log of cibadmin attached > > (with the two nodes of the cluster renamed foo and bar). > > > > [EMAIL PROTECTED] crm_mon -1 > > > > > > ============ > > Last updated: Thu Jan 26 09:13:02 2006 > > Current DC: foo.inria.fr (5e9d32e2-7806-496f-bce3-a45124c608a4) > > 1 Nodes configured. > > 1 Resources configured. > > ============ > > > > Node: foo.inria.fr (5e9d32e2-7806-496f-bce3-a45124c608a4): online > > > > Resource Group: dns-sop1 > > IPaddr_dns-sop1 (heartbeat::ocf:IPaddr): foo.inria.fr > > named_dns-sop1 (heartbeat::ocf:named): foo.inria.fr > > [EMAIL PROTECTED] > > > > ## I started thus heartbeat only on node foo for this test. > > > > [EMAIL PROTECTED] crm_verify -X /tmp/cib.xml -V > > crm_verify[1959]: 2006/01/26_08:48:50 WARN: mask(stages.c:stage0): > > We do > > not have quorum - fencing and resource management disabled > > [EMAIL PROTECTED] > > > > ## I don't understand this warning. Something to worry about? > > I would think so don't you? > > > > > [EMAIL PROTECTED] cibadmin --id foo-bar -U -VVVVVV -x /tmp/cib.xml >& > > /tmp/cibadmin.out > > > > ## /tmp/cib.xml defines the value of transition_idle_timeout to be > > 1515 > > > > [EMAIL PROTECTED] cibadmin -Q|grep transition_idle_timeout > > <nvpair id="transition_idle_timeout" > > name="transition_idle_ti > > meout" value="200s"/> > > [EMAIL PROTECTED] > > > > ## No change :-( > > /me bangs head against desk (try CVS in a few minutes I found a > really dumb bug) > > otherwise, try: cibadmin -M -o crm_config -X "<nvpair id= > \"transition_idle_timeout\" value=\"5m\"/>" > > -- > Andrew Beekhof > > "No means no, and no means yes, and everything in between and all the > rest" - TISM > > _______________________________________________________ > Linux-HA-Dev: [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev > Home Page: http://linux-ha.org/ > _______________________________________________________ Linux-HA-Dev: [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev Home Page: http://linux-ha.org/
