AW: [Linux-HA] Heartbeat v2 stickiness, score and more
> > Hello, > > I spent the whole afternoon to search for a good heartbeat v2 > > documentation, but it looks like this is somehow difficult. Maybe > > someone in here can help me? > > The only good doc is Andrews Configuration_Explained. See: > http://www.clusterlabs.org/mediawiki/images/6/61/Colocation_Ex > plained.pdf Thank you Michael nice link! http://www.clusterlabs.org/mediawiki/images/f/fb/Configuration_Explained.pdf > > On the other hand, if you understand German, I could suggest > you a very good book from O'Reilly ;-) > Yes we will order this book. > > Anyway I have a short question about "stickiness". I only > know about > > sun cluster but I have to build up knowledge about > heartbeat cluster > > since we are running two debian heartbeat clusters now. > > Be sure to get really the latest packages from Linux-HA. The > default debian packages are a little behind. There exist a > lot of improvements and bug fixes in the mean time. > At the moment we use the stable debian lenny packages. Do you know any good debian repository for the latest HA packages? > > Those two are failover clusters providing web services, nagios and > > vserver virtual hosts. Let's say resource_group_a is > running on node1 > > and resource_group_b on node2. If I reboot node2 > resource_group_b will > > switch to node1. But if node2 is up again resource_group_b > will switch > > back to node2. That is what I don't want the cluster to do. No > > switchback... How can I do that? > > Set the resource-stickness of the group_b lets say to 100 > points. Then the group will stay on the node it runs on until > the difference in score calculation due to other constraints > is more than 100 points. > Works perfect! Thank you! > > And which command is used to switch one resource group to > another node > > (not marking any node as "standby")? > > > crm_resource -M -r -H > > unmigrate with > > crm_resource -U r > Also worked perfect! Thank you Michael. Best regards Florian > > Thank you for your help. > > > > Best regards Florian > > ___ > > Linux-HA mailing list > > Linux-HA@lists.linux-ha.org > > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > See also: http://linux-ha.org/ReportingProblems > > -- > Dr. Michael Schwartzkopff > MultiNET Services GmbH > Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany > Tel: +49 - 89 - 45 69 11 0 > Fax: +49 - 89 - 45 69 11 21 > mob: +49 - 174 - 343 28 75 > > mail: mi...@multinet.de > web: www.multinet.de > > Sitz der Gesellschaft: 85630 Grasbrunn > Registergericht: Amtsgericht München HRB 114375 > Geschäftsführer: Günter Jurgeneit, Hubert Martens > > --- > > PGP Fingerprint: F919 3919 FF12 ED5A 2801 DEA6 AA77 57A4 EDD8 979B > Skype: misch42 > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Heartbeat v2 stickiness, score and more
florian.engelm...@bt.com wrote: > Hello, > I spent the whole afternoon to search for a good heartbeat v2 > documentation, but it looks like this is somehow difficult. Maybe > someone in here can help me? > > Anyway I have a short question about "stickiness". I only know about sun > cluster but I have to build up knowledge about heartbeat cluster since > we are running two debian heartbeat clusters now. > Those two are failover clusters providing web services, nagios and > vserver virtual hosts. Let's say resource_group_a is running on node1 > and resource_group_b on node2. If I reboot node2 resource_group_b will > switch to node1. But if node2 is up again resource_group_b will switch > back to node2. That is what I don't want the cluster to do. No > switchback... How can I do that? crm_attribute --type crm_config --attr-name default-resource-stickiness --attr-value INFINITY Old versions might need "_" instead of "-" in the attr-name. If yours does, take that as a hint that you should upgrade your cluster software. > And which command is used to switch one resource group to another node > (not marking any node as "standby")? crm_resource Regards Dominik ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] 1 ip resource fail between master master mysql instance help
I had just about the same use-case as you; I posted my configuration on the pacemaker list: http://www.mail-archive.com/pacema...@oss.clusterlabs.org/msg00356.html Good luck! Neil Katin Martin Suehowicz wrote: Thank You, I am almost there. How do have both servers running mysqld? And have a collocation resource between the 2. ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] rule move resource if mysql status fails
Hello, Can someone tell me if it's possible to move a resource off a server if /etc/init.d/mysql status fails. Or perhaps if a process is not running. I don't want halinux to manage mysql. I see in docs that there is a way to say move a resource if a fibre card does not exist or if it has not enough cpus. Is there a list of rules somewhere? Thank you for any help you provide! Martin ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
RE: [Linux-HA] 1 ip resource fail between master master mysql instance help
Thank You, I am almost there. How do have both servers running mysqld? And have a collocation resource between the 2. -Original Message- From: linux-ha-boun...@lists.linux-ha.org [mailto:linux-ha-boun...@lists.linux-ha.org] On Behalf Of Andrew Beekhof Sent: Wednesday, April 01, 2009 8:00 AM To: General Linux-HA mailing list Subject: Re: [Linux-HA] 1 ip resource fail between master master mysql instance help add a colocation constraint between the ip and mysql. On Wed, Apr 1, 2009 at 08:43, Martin Suehowicz wrote: > I am setting up a master/master mysql instance with 2 arrays. I would > like to setup a address to fail between the two of them. > I can setup a ip resource ok and I can get to fail over ok. I do have 2 > problems. > > 1. I cannot figure out how to fail the ip resource if mysql goes down. > Also since it is master master I need both mysql resources running. > 2. I also want to disable auto fail back. "auto_failback off" in > /etc/ha.d/ha.cf does not work. Help! Here is my cib.xml: > > > > ignore_dtd="false" ccm_transition="5" num_peers="1" > cib_feature_revision="2.0" crm_feature_set="2.0" > dc_uuid="480621cf-cb29-4ac4-887f-8572da0d92d7" epoch="7" num_updates="2" > cib-last-written="Sat Oct 11 09:13:11 2008"> > > > > > name="dc-version" value="2.1.4-node: > 74d736e8f667fae1e4a904ccd13ad6553c5c010c"/> > > > > > uname="qsql01.lax.rubi" type="normal"/> > uname="qsql02.lax.rubi" type="normal"/> > > > provider="heartbeat"> > > > > > > > > > > > > > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] Watchdog and apphbd question
Can someone please explain to me how I can get my hardware watchdog to monitor processes and reboot if a process hangs or is killed in a non graceful manner? I have heartbeat running as it should. I have a hardware watchdog set up and I have confirmed that it is working properly. However, I would like to make it possible so if something happens to a running daemon (like heartbeat as a whole), it will trigger watchdog to stop resetting the WDT, therefor rebooting the machine. I might be completely off when I mention this, but I was under the impression that the apphbd plugin is supposed to do this. If so, how do I set this up? I have added "watchdog /dev/watchdog" into the ha.cf file, and performed a killall -9 heartbeat. But the machine does not reboot. I am also aware that I can set up watchdog so if the load average goes above a certain level, a reboot will occur. But is there another, more efficient way to reboot the system if a process is killed off unexpectedly or if the process is hung? I can only assume that the load average will naturally go up if this were the case, so perhaps just setting the load average setting may do just fine. Or maybe I've had too much coffee and I've been thinking too fast and deep about this, and have made it more complicated than it already is... %-| Thanks for your help! -- View this message in context: http://www.nabble.com/Watchdog-and-apphbd-question-tp22834997p22834997.html Sent from the Linux-HA mailing list archive at Nabble.com. ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Adding a Hostname to a Clustered Resource
Hi Dejan Sorry I should have explained myself better i have a 2 node DRBD cluster, doing file and printer sharing, i have been able to get an IP Address to move with the resources, and was hoping to have the virtual server name move along with the IP address, and if possible I would like this server name to automatically register with DNS (damn those clowns!!) I am pretty sure that I had this working with an older version (pre pacemaker) but at the time I had used the GUI to configure and I have no idea on how to do this using the CRM thanks again Jay On Wed, 2009-04-01 at 14:50 +0200, Dejan Muhamedagic wrote: > Hi, > > On Wed, Apr 01, 2009 at 11:01:50AM +0100, Jason Fitzpatrick wrote: > > Hi all > > > > having a bit of a tough time trying to find the setting for hostnames within > > CRM > > setting for hostnames? Which/where? If you want to see nodes with > crm, please update, it got implemented recently. > > > my current CRM\show config is as follows > > > > primitive DRBD_Disk ocf:heartbeat:drbd \ > > params drbd_resource="Storage1" \ > > op monitor interval="59s" role="Master" timeout="30s" \ > > op monitor interval="60s" role="Slave" timeout="30s" \ > > meta target-role="started" > > primitive Storage ocf:heartbeat:Filesystem \ > > params fstype="xfs" directory="/Storage" device="/dev/drbd0" \ > > meta target-role="started" > > primitive Cluster_IP_Address ocf:heartbeat:IPaddr2 \ > > params ip="10.7.4.165" broadcast="10.7.4.255" nic="eth2" > > cidr_netmask="24" \ > > op monitor interval="21s" timeout="5s" > > primitive CUPS lsb:cups \ > > operations $id="CUPS-operations" \ > > op monitor interval="15" timeout="5" \ > > meta target-role="started" > > primitive Samba lsb:smb \ > > op monitor interval="21s" timeout="5s" > > group Cluster_Resources Storage Cluster_IP_Address CUPS Samba > > ms DRBD_Resource DRBD_Disk \ > > meta clone-max="2" notify="true" globally-unique="false" > > target-role="Started" > > colocation Cluster_Resource-Same-Node-as-DRBD_Resource inf: > > Cluster_Resources DRBD_Resource:Master > > order DRBD_Resource-Before-Cluster_Resource inf: DRBD_Resource:promote > > Cluster_Resources:start > > property $id="cib-bootstrap-options" \ > > dc-version="1.0.2-c02b459053bfa44d509a2a0e0247b291d93662b7" \ > > last-lrm-refresh="1236618834" > > > > and I would realy like for this cluster to automatically register its name > > in DNS because some clown in here keeps deleting the A Record for this > > server! > > If you can't keep the clown away (though this is an unusual > business you're talking about), then you'll have to write an RA > which would undo the said clown's operation. > > Thanks, > > Dejan > > > Thanks a mill > > > > Jason > > ___ > > Linux-HA mailing list > > Linux-HA@lists.linux-ha.org > > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > See also: http://linux-ha.org/ReportingProblems > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Heartbeat v2 stickiness, score and more
Am Mittwoch, 1. April 2009 20:29:57 schrieb Michael Schwartzkopff: > Am Mittwoch, 1. April 2009 17:55:39 schrieb florian.engelm...@bt.com: > > Hello, > > I spent the whole afternoon to search for a good heartbeat v2 > > documentation, but it looks like this is somehow difficult. Maybe > > someone in here can help me? > > The only good doc is Andrews Configuration_Explained. See: > http://www.clusterlabs.org/mediawiki/images/6/61/Colocation_Explained.pdf Sorry, I mean of course: http://www.clusterlabs.org/mediawiki/images/f/fb/Configuration_Explained.pdf (...) -- Dr. Michael Schwartzkopff MultiNET Services GmbH Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany Tel: +49 - 89 - 45 69 11 0 Fax: +49 - 89 - 45 69 11 21 mob: +49 - 174 - 343 28 75 mail: mi...@multinet.de web: www.multinet.de Sitz der Gesellschaft: 85630 Grasbrunn Registergericht: Amtsgericht München HRB 114375 Geschäftsführer: Günter Jurgeneit, Hubert Martens --- PGP Fingerprint: F919 3919 FF12 ED5A 2801 DEA6 AA77 57A4 EDD8 979B Skype: misch42 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Heartbeat v2 stickiness, score and more
Am Mittwoch, 1. April 2009 17:55:39 schrieb florian.engelm...@bt.com: > Hello, > I spent the whole afternoon to search for a good heartbeat v2 > documentation, but it looks like this is somehow difficult. Maybe > someone in here can help me? The only good doc is Andrews Configuration_Explained. See: http://www.clusterlabs.org/mediawiki/images/6/61/Colocation_Explained.pdf On the other hand, if you understand German, I could suggest you a very good book from O'Reilly ;-) > Anyway I have a short question about "stickiness". I only know about sun > cluster but I have to build up knowledge about heartbeat cluster since > we are running two debian heartbeat clusters now. Be sure to get really the latest packages from Linux-HA. The default debian packages are a little behind. There exist a lot of improvements and bug fixes in the mean time. > Those two are failover clusters providing web services, nagios and > vserver virtual hosts. Let's say resource_group_a is running on node1 > and resource_group_b on node2. If I reboot node2 resource_group_b will > switch to node1. But if node2 is up again resource_group_b will switch > back to node2. That is what I don't want the cluster to do. No > switchback... How can I do that? Set the resource-stickness of the group_b lets say to 100 points. Then the group will stay on the node it runs on until the difference in score calculation due to other constraints is more than 100 points. > And which command is used to switch one resource group to another node > (not marking any node as "standby")? crm_resource -M -r -H unmigrate with crm_resource -U r > Thank you for your help. > > Best regards Florian > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems -- Dr. Michael Schwartzkopff MultiNET Services GmbH Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany Tel: +49 - 89 - 45 69 11 0 Fax: +49 - 89 - 45 69 11 21 mob: +49 - 174 - 343 28 75 mail: mi...@multinet.de web: www.multinet.de Sitz der Gesellschaft: 85630 Grasbrunn Registergericht: Amtsgericht München HRB 114375 Geschäftsführer: Günter Jurgeneit, Hubert Martens --- PGP Fingerprint: F919 3919 FF12 ED5A 2801 DEA6 AA77 57A4 EDD8 979B Skype: misch42 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Determining whether CIB is accessible
On Wed, Mar 18, 2009 at 3:44 AM, Andrew Beekhof wrote: > > On Tue, Mar 17, 2009 at 16:10, Nicholas Dronen wrote: > > Hi: > > > > I'm writing scripts to dynamically add, remove, manage, and unmanage > > resources from the CIB. We already have scripts that start and stop > > heartbeat. What I'm doing is updating our code so we can use V2. To > > add resources to a V1 configuration, you simply added the resource > > information to /etc/ha.d/haresources; if heartbeat was already > > running, you told heartbeat to reload the configuration, if not you > > could start heartbeat. With a V2 configuration, however, it seems > > like heartbeat has to be running before you can change the CIB. This > > implies the need for a way to determine whether the CIB is available. > > One way I see is to run 'crm resource status 2>/dev/null' and look for > > a 0 return code, since it returns 1 if heartbeat isn't running > > locally. (I should note that 'crm configure show' returns 0 even if > > it can't connect to heartbeat.) > > Create a bug for that, it shouldn't. OK. How do I do that? > > Is this a reasonable thing to do? > > Anything to be careful of when using this approach? > > Sounds fine. > But would this be an issue if the tools you're using accurately > reported failure when they can't perform the change? That would work as well. Should I open a bug for that, too? (Sorry for the late response.) Regards, Nick ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] Heartbeat v2 stickiness, score and more
Hello, I spent the whole afternoon to search for a good heartbeat v2 documentation, but it looks like this is somehow difficult. Maybe someone in here can help me? Anyway I have a short question about "stickiness". I only know about sun cluster but I have to build up knowledge about heartbeat cluster since we are running two debian heartbeat clusters now. Those two are failover clusters providing web services, nagios and vserver virtual hosts. Let's say resource_group_a is running on node1 and resource_group_b on node2. If I reboot node2 resource_group_b will switch to node1. But if node2 is up again resource_group_b will switch back to node2. That is what I don't want the cluster to do. No switchback... How can I do that? And which command is used to switch one resource group to another node (not marking any node as "standby")? Thank you for your help. Best regards Florian ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Programmatic interface to CRM?
I am delighted to hear there is subscription mechanism! I apologize if I have not read through the documentation thoroughly. Could you please point me to the documentation on CIB events and the subscription mechanism? What is the protocol by which events are conveyed to subscribers? Is there a hello world example of how to subscribe to CIB events? Tx, /Uma - Original Message - From: Andrew Beekhof To: General Linux-HA mailing list Sent: Wed, 1 Apr 2009 06:51:21 + (UTC) Subject: Re: [Linux-HA] Programmatic interface to CRM? On Tue, Mar 31, 2009 at 22:36, wrote: > Is there a programmatic interface to CRM available, by which > an application could query CIB and perhaps be notified (based > on callbacks or some such) about events such as transitions? you can subscribe to cib events. crm_mon does this to generate snmp and email alerts for resource actions. eventually we'll also support snmp queries and alert filters. > Alternatively, is there a way to have CRM trigger a script when > critical events occur? We have a central clearing house of > events from where events are ent out as notifications as > Syslog, SNMP, Email or Jabber. > > I would like to feed CRM events into this event clearing house. > > CRM documentations hints that such a mechanism is not yet available: > > TODO: Should traps/events be triggered from inside the different > >components themselves, or might it be a good idea to allow >clients to > "subscribe" to certain parts of the CIB and be >notified if a change > occurs? This would be helpful for >SNMP/CIM traps. (The later would be > somewhat alike FailSafe's cdbd) > > If neither is possible, could someone suggest an approximation > to this functionality? > > Regards, > > /Uma > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] 1 ip resource fail between master master mysql instance help
add a colocation constraint between the ip and mysql. On Wed, Apr 1, 2009 at 08:43, Martin Suehowicz wrote: > I am setting up a master/master mysql instance with 2 arrays. I would > like to setup a address to fail between the two of them. > I can setup a ip resource ok and I can get to fail over ok. I do have 2 > problems. > > 1. I cannot figure out how to fail the ip resource if mysql goes down. > Also since it is master master I need both mysql resources running. > 2. I also want to disable auto fail back. "auto_failback off" in > /etc/ha.d/ha.cf does not work. Help! Here is my cib.xml: > > > > ignore_dtd="false" ccm_transition="5" num_peers="1" > cib_feature_revision="2.0" crm_feature_set="2.0" > dc_uuid="480621cf-cb29-4ac4-887f-8572da0d92d7" epoch="7" num_updates="2" > cib-last-written="Sat Oct 11 09:13:11 2008"> > > > > > name="dc-version" value="2.1.4-node: > 74d736e8f667fae1e4a904ccd13ad6553c5c010c"/> > > > > > uname="qsql01.lax.rubi" type="normal"/> > uname="qsql02.lax.rubi" type="normal"/> > > > provider="heartbeat"> > > > > > > > > > > > > > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] ucast on multiple IPs
Hi, On Sun, Mar 29, 2009 at 11:11:50PM +0200, m...@bortal.de wrote: > Hello List, > > i have a linux cluster with 2 nodes. Each node has two network interfaces. > Can i now do ucasts on all of the interfaces. So ha.cf would look like: > > ucast eth0 > ucast eth1 > > Therefore BOTH ucasts have to fail before a takeover takes place? Does this > work? Right. > Due to hardware limitation, i can not use a direkt crossed link between a > 3rd nic. What kind of hw limitation did you encounter? Thanks, Dejan > Cheers, > Mario > > > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Adding a Hostname to a Clustered Resource
Hi, On Wed, Apr 01, 2009 at 11:01:50AM +0100, Jason Fitzpatrick wrote: > Hi all > > having a bit of a tough time trying to find the setting for hostnames within > CRM setting for hostnames? Which/where? If you want to see nodes with crm, please update, it got implemented recently. > my current CRM\show config is as follows > > primitive DRBD_Disk ocf:heartbeat:drbd \ > params drbd_resource="Storage1" \ > op monitor interval="59s" role="Master" timeout="30s" \ > op monitor interval="60s" role="Slave" timeout="30s" \ > meta target-role="started" > primitive Storage ocf:heartbeat:Filesystem \ > params fstype="xfs" directory="/Storage" device="/dev/drbd0" \ > meta target-role="started" > primitive Cluster_IP_Address ocf:heartbeat:IPaddr2 \ > params ip="10.7.4.165" broadcast="10.7.4.255" nic="eth2" > cidr_netmask="24" \ > op monitor interval="21s" timeout="5s" > primitive CUPS lsb:cups \ > operations $id="CUPS-operations" \ > op monitor interval="15" timeout="5" \ > meta target-role="started" > primitive Samba lsb:smb \ > op monitor interval="21s" timeout="5s" > group Cluster_Resources Storage Cluster_IP_Address CUPS Samba > ms DRBD_Resource DRBD_Disk \ > meta clone-max="2" notify="true" globally-unique="false" > target-role="Started" > colocation Cluster_Resource-Same-Node-as-DRBD_Resource inf: > Cluster_Resources DRBD_Resource:Master > order DRBD_Resource-Before-Cluster_Resource inf: DRBD_Resource:promote > Cluster_Resources:start > property $id="cib-bootstrap-options" \ > dc-version="1.0.2-c02b459053bfa44d509a2a0e0247b291d93662b7" \ > last-lrm-refresh="1236618834" > > and I would realy like for this cluster to automatically register its name > in DNS because some clown in here keeps deleting the A Record for this > server! If you can't keep the clown away (though this is an unusual business you're talking about), then you'll have to write an RA which would undo the said clown's operation. Thanks, Dejan > Thanks a mill > > Jason > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] Adding a Hostname to a Clustered Resource
Hi all having a bit of a tough time trying to find the setting for hostnames within CRM my current CRM\show config is as follows primitive DRBD_Disk ocf:heartbeat:drbd \ params drbd_resource="Storage1" \ op monitor interval="59s" role="Master" timeout="30s" \ op monitor interval="60s" role="Slave" timeout="30s" \ meta target-role="started" primitive Storage ocf:heartbeat:Filesystem \ params fstype="xfs" directory="/Storage" device="/dev/drbd0" \ meta target-role="started" primitive Cluster_IP_Address ocf:heartbeat:IPaddr2 \ params ip="10.7.4.165" broadcast="10.7.4.255" nic="eth2" cidr_netmask="24" \ op monitor interval="21s" timeout="5s" primitive CUPS lsb:cups \ operations $id="CUPS-operations" \ op monitor interval="15" timeout="5" \ meta target-role="started" primitive Samba lsb:smb \ op monitor interval="21s" timeout="5s" group Cluster_Resources Storage Cluster_IP_Address CUPS Samba ms DRBD_Resource DRBD_Disk \ meta clone-max="2" notify="true" globally-unique="false" target-role="Started" colocation Cluster_Resource-Same-Node-as-DRBD_Resource inf: Cluster_Resources DRBD_Resource:Master order DRBD_Resource-Before-Cluster_Resource inf: DRBD_Resource:promote Cluster_Resources:start property $id="cib-bootstrap-options" \ dc-version="1.0.2-c02b459053bfa44d509a2a0e0247b291d93662b7" \ last-lrm-refresh="1236618834" and I would realy like for this cluster to automatically register its name in DNS because some clown in here keeps deleting the A Record for this server! Thanks a mill Jason ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] pingd/pacemaker
Am Mittwoch, 1. April 2009 08:52:37 schrieb Dominik Klein: > > I know. But this attrbiut does not exist in my setup. pacemaker verison > > 1.0.1-1. Is this a feature of 1.0.2? > > 1.0.1 is 4 months old. The RA was updated with those features 3 months > ago. So basically, yes. You could still update the single RA from the > mercurial repository though. > > Regards > Dominik Thanks for the info. The development speed is too fast to catch up with all changes ;-) Greetings, -- Dr. Michael Schwartzkopff MultiNET Services GmbH Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany Tel: +49 - 89 - 45 69 11 0 Fax: +49 - 89 - 45 69 11 21 mob: +49 - 174 - 343 28 75 mail: mi...@multinet.de web: www.multinet.de Sitz der Gesellschaft: 85630 Grasbrunn Registergericht: Amtsgericht München HRB 114375 Geschäftsführer: Günter Jurgeneit, Hubert Martens --- PGP Fingerprint: F919 3919 FF12 ED5A 2801 DEA6 AA77 57A4 EDD8 979B Skype: misch42 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] showscores.sh for pacemaker 1.0.2
So here's an update. Michael Schwartzkopf pointed out a bug regarding groups. That has been fixed now and the appropriate values should be shown. Thanks! There's not been a lot of feedback, is it because nobody uses the script or does it just work for you? Regards Dominik Dominik Klein wrote: > Hi > > I made the necessary changes to the showscores script to work with > pacemaker 1.0.2. > > Please test and report problems. Has been reported to work by some > people and should go into the repository soon. Still, I'd like more > people to test and confirm. > > Important changes: > * correctly fetch stickiness and migration-threshold for complex > resources (master and clone) > * adjust column-width according to the length of resources' and nodes' names > > Regards > Dominik showscores.sh Description: Bourne shell script ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] 1 ip resource fail between master master mysql instance help
I am setting up a master/master mysql instance with 2 arrays. I would like to setup a address to fail between the two of them. I can setup a ip resource ok and I can get to fail over ok. I do have 2 problems. 1. I cannot figure out how to fail the ip resource if mysql goes down. Also since it is master master I need both mysql resources running. 2. I also want to disable auto fail back. "auto_failback off" in /etc/ha.d/ha.cf does not work. Help! Here is my cib.xml: ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Programmatic interface to CRM?
On Tue, Mar 31, 2009 at 22:36, wrote: > Is there a programmatic interface to CRM available, by which > an application could query CIB and perhaps be notified (based > on callbacks or some such) about events such as transitions? you can subscribe to cib events. crm_mon does this to generate snmp and email alerts for resource actions. eventually we'll also support snmp queries and alert filters. > Alternatively, is there a way to have CRM trigger a script when > critical events occur? We have a central clearing house of > events from where events are ent out as notifications as > Syslog, SNMP, Email or Jabber. > > I would like to feed CRM events into this event clearing house. > > CRM documentations hints that such a mechanism is not yet available: > > TODO: Should traps/events be triggered from inside the different > > components themselves, or might it be a good idea to allow > clients to > "subscribe" to certain parts of the CIB and be > notified if a change > occurs? This would be helpful for > SNMP/CIM traps. (The later would be > somewhat alike FailSafe's cdbd) > > If neither is possible, could someone suggest an approximation > to this functionality? > > Regards, > > /Uma > ___ > Linux-HA mailing list > Linux-HA@lists.linux-ha.org > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems > ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] 1 ip resource fail between master master mysql instance help
I am setting up a master/master mysql instance with 2 arrays. I would like to setup a address to fail between the two of them. I can setup a ip resource ok and I can get to fail over ok. I do have 2 problems. 1. I cannot figure out how to fail the ip resource if mysql goes down. Also since it is master master I need both mysql resources running. 2. I also want to disable auto fail back. "auto_failback off" in /etc/ha.d/ha.cf does not work. Help! Here is my cib.xml: ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
