Re: [Pacemaker] symmetric anti-collocation
On Fri, Nov 12, 2010 at 5:27 PM, Alan Jones falanclus...@gmail.com wrote: On Thu, Nov 11, 2010 at 11:31 PM, Andrew Beekhof and...@beekhof.net wrote: colocation X-Y -2: X Y colocation Y-X -2: Y X the second one is implied by the first and is therefore redundant If only that were true! It is. I know exactly how my code works in this regard. More than likely a score of -2 is simply too low to have any effect. What happens with the first rule is that other constraints that force Y to a node will evict X but not the other way around. What I'm doing is to first apply a slight preference for each resource to each node: location X-nodeA X 1: nodeA location Y-nodeB Y 1: nodeB And then impose absolute constraints that come from the outside environment. In the particular case that has a problem, the constraint looks like: location X-not-nodeA X -inf: nodeA The behavior I expected was for X to be placed on nodeB and Y to anti-colocate onto nodeA because our colocation rule is stronger than the node preference rule. What happens instead is that both X and Y run on nodeB. The similar constraint on Y (by itself) does work: location Y-not-nodeB Y -inf: nodeB and results in Y running on nodeA and X running on nodeB. This is the case whether I have one colocation rule or two, i.e. the second colocation rule is ignored. Looking at the code, I think the solution would be to short-circuit the recursion when you can only run on one node due to -inf rules rather than on a loop. Obviously, it would not be a simple change and needs some thought. If you have any other suggestions let me know. Alan ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] Project updates
On Fri, Nov 12, 2010 at 5:49 PM, Vadym Chepkov vchep...@gmail.com wrote: On Fri, Nov 12, 2010 at 9:32 AM, Andrew Beekhof and...@beekhof.net wrote: For those that aren't using RSS readers, I wanted to draw people's attention to a couple of updates that went out today. Nothing dramatic, just a new 1.0 release (and back-annoucement for some from 1.1): Perhaps something was forgotten in the excitement? :) $ hg diff diff -r 99f5a1e61667 GNUmakefile --- a/GNUmakefile Fri Nov 12 09:12:32 2010 +0100 +++ b/GNUmakefile Fri Nov 12 11:47:28 2010 -0500 @@ -26,7 +26,7 @@ TARFILE = $(distdir).tar.bz2 DIST_ARCHIVES = $(TARFILE) -LAST_RELEASE = Pacemaker-1.0.9.1 +LAST_RELEASE = Pacemaker-1.0.10 thats only used for calculating a new changelog :-) STABLE_SERIES = stable-1.0 RPM_ROOT = $(shell pwd) diff -r 99f5a1e61667 configure.ac --- a/configure.ac Fri Nov 12 09:12:32 2010 +0100 +++ b/configure.ac Fri Nov 12 11:47:28 2010 -0500 @@ -19,7 +19,7 @@ dnl checks for library functions dnl checks for system services -AC_INIT(pacemaker, 1.0.9, pacemaker@oss.clusterlabs.org) +AC_INIT(pacemaker, 1.0.10, pacemaker@oss.clusterlabs.org) thats kinda annoying but not crucial. thanks for pointing it out CRM_DTD_VERSION=1.0 PKG_FEATURES= ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] symmetric anti-collocation
On Sat, Nov 13, 2010 at 3:20 AM, Andrew Beekhof and...@beekhof.net wrote: On Fri, Nov 12, 2010 at 5:27 PM, Alan Jones falanclus...@gmail.com wrote: On Thu, Nov 11, 2010 at 11:31 PM, Andrew Beekhof and...@beekhof.net wrote: colocation X-Y -2: X Y colocation Y-X -2: Y X the second one is implied by the first and is therefore redundant If only that were true! It is. I know exactly how my code works in this regard. More than likely a score of -2 is simply too low to have any effect. I have tried larger values. If you know of a value that *should* work, please share it. I wish to be able to predict the behavior, so arbitrary values are not acceptable. It is either large enough to overcome competing values or not. In this configuration I've set the default resource stickyness to zero and set a node location rule per resource to one. It is clear from the logs that the -2 is not being applied. I also tried unique large numbers to see if I could trace their propagation, e.g. -1965, with no effect. If you have time to look at this, I will have to create sanitized versions of logs and configuration, etc. Let me know. Alan ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] 2 node failover cluster + MySQL Master-Master replica setup
That's what I said - I didn't see it either. but if you you check the current RA: What do you think about this: http://www.lathiat.net/files/MySQL%20-%20DRBD%20%20Pacemaker.pdf I can't see if this is a real M-M or M-S setup. TIA, Ruzsi ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] symmetric anti-collocation
On Sat, Nov 13, 2010 at 5:09 PM, Alan Jones falanclus...@gmail.com wrote: On Sat, Nov 13, 2010 at 3:20 AM, Andrew Beekhof and...@beekhof.net wrote: On Fri, Nov 12, 2010 at 5:27 PM, Alan Jones falanclus...@gmail.com wrote: On Thu, Nov 11, 2010 at 11:31 PM, Andrew Beekhof and...@beekhof.net wrote: colocation X-Y -2: X Y colocation Y-X -2: Y X the second one is implied by the first and is therefore redundant If only that were true! It is. I know exactly how my code works in this regard. More than likely a score of -2 is simply too low to have any effect. I have tried larger values. If you know of a value that *should* work, please share it. INFINITY I wish to be able to predict the behavior, so arbitrary values are not acceptable. It is either large enough to overcome competing values or not. In this configuration I've set the default resource stickyness to zero and set a node location rule per resource to one. It is clear from the logs that the -2 is not being applied. I also tried unique large numbers to see if I could trace their propagation, e.g. -1965, with no effect. If you have time to look at this, I will have to create sanitized versions of logs and configuration, etc. Let me know. Alan ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
[Pacemaker] (no subject)
pLunch this week?brbrbr/p pSent from Yahoo! Mail on Android/p ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] symmetric anti-collocation
I have tried larger values. If you know of a value that *should* work, please share it. INFINITY My understanding is that a colocation score of minus infinity will prevent the resources from running on the same node, which in my configuration would result in a loss of availability. The goal is a value that will overrule the node location preferences of one, allow for both resources to run on the same node, and provide for anti-colocation to evict either resource based on the constraints of the other. I can't claim to understand your code, but I've done sufficient experiments to understand the code's behavior in 1.0.9.1. There is simply no combination of colocation rules and values that will work in both directions. I can get the placement of Y to evict X -- or -- X to evict Y; but not both. Alan ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Re: [Pacemaker] start filesystem like this is right?
start resource steps step(1) crm configure primitive?vol_mpath0 ocf:heartbeat:Filesystem meta target-role=stopped params device=/dev/mapper/mpath0 directory=/mnt/mapper/mpath0 fstype='lustre' op start timeout=300s? op stop timeout=120s op monitor timeout=120s interval=60s op notify timeout=60s step(2)crm resource reprobe step(3) crm configure location vol_mpath0_location_manage?datavol_mpath0 rule -inf: not_defined pingd_manage or pingd_manage lte 0 crm configure location vol_mpath0_location_data?datavol_mpath0 rule -inf: not_defined pingd_data or pingd_data lte 0 why do you have 2 location constraints? where is the definitions for pingd_data and pingd_manage? because we have two network. manage network is ethernet data network is ib. the definitions of pingd crm configure primitive pingd_data ocf:pacemaker:ping meta target-role=stopped params name=pingd_data op start timeout=100s op stop timeout=100s op monitor interval=90s timeout=100s; crm_resource -p host_list -r pingd_data -v IP_list crm configure clone pingd_data_net pingd_data meta globally-unique=falsetarget-role=stopped crm resource start pingd_data crm configure primitive pingd_manage ocf:pacemaker:ping meta target-role=stopped params name=pingd_manage op start timeout=90s op stop timeout=100s op monitor interval=90s timeout=100s crm_resource -p host_list -r pingd_manage -v IP_list crm configure clone pingd_manage_net pingd_manage meta globally-unique=false crm resource start pingd_manage step(4) .crm resource start vol_mpath0 delete resource steps .step(1) crm resource stop vol_mpath0 step(2) crm resource cleanup vol_mpath0 step(3) crm configure delete vol_mpath0 above?is my steps? is it right? I repeat these steps for several times. at begin it works well. after 5 or 6 times the reosurce could not start .I use crm resource start vol_mpath0 again no use. Could be that your ping nodes are down? the node is ok. and do you know the way to check pingd? and I found the cluster is something wrong.I check the log I think node could not get the situation from crm my pacemaker package?are ? ? pacemaker-1.0.8-6.1.el5 ? ? pacemaker-libs-devel-1.0.8-6.1.el5 ? ? pacemaker-libs-1.0.8-6.1.el5 ? ? openais packages?are ? ? openaislib-devel-1.1.0-1.el5 ? ? openais-1.1.0-1.el5 ? ? openaislib-1.1.0-1.el5 ??? corosync packages are ? ? corosync-1.2.2-1.1.el5 ? ? corosynclib-devel-1.2.2-1.1.el5 ? ? corosynclib-1.2.2-1.1.el5 ? ? who know why thanks a lot ___ Pacemaker mailing list: Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker -- Message: 4 Date: Fri, 12 Nov 2010 13:07:56 -0500 From: Vadym Chepkov vchep...@gmail.com To: The Pacemaker cluster resource manager pacemaker@oss.clusterlabs.org Subject: Re: [Pacemaker] 2 node failover cluster + MySQL Master-Master replica setup Message-ID: 12802784-1613-4ef6-81ff-7b91f2b3b...@gmail.com Content-Type: text/plain; charset=us-ascii On Nov 12, 2010, at 11:14 AM, Ruzsinszky Attila wrote: Hi, And now (officially) RHCS can also use Pacemaker http://theclusterguy.clusterlabs.org/post/1551292286 Nice. Yeah, like I said, Master-Master and Pacemaker without a proper resource agent will cause issues. Yes. big problems. Now let me explain this, a 2-node Multi-Master MySQL setup means setting up every node as both Master and Slave, node 1's Master replicates asynchronously to node 2's Slave and node 2's Master replicates asynchronously to node 1's Slave. The replication channels between the two are not redundant, nor do they recover from failure automatically and you have to manually set the auto-increment-increment and auto-increment-offset so that you don't have primary key collisions. Clear. each server. Looking at how DRBD handles these kinds of things is one way to go about it, but ... it's a huge task and there are a lot of things that can go terribly wrong. :-( So again, for the third time, the problem is not the Multi-Master setup, nor it is Pacemaker, it's just a very specific use case for which a resource agent wasn't written. OK. So now almost the only one possibilities is DRBD+MySQL? I am pretty sure Linbit announced mysql RA with replication capabilities. Haven't seen documentation though. Vadym -- ___ Pacemaker mailing list Pacemaker@oss.clusterlabs.org http://oss.clusterlabs.org/mailman/listinfo/pacemaker End of Pacemaker Digest, Vol 36, Issue 47 * ___ Pacemaker mailing list: