Re: [Linux-ha-dev] [Patch 3]Change avoiding the stop error of the mysql resource agent.
Hi Hideo-san! On 09/12/2011 03:19 AM, renayama19661...@ybb.ne.jp wrote: Hi Raoul, How about the modified patch of this place? unfortunatly, i'm currently a little swamped. i'll look into that asap. please bear with me a little longer! thanks, raoul ___ Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev Home Page: http://linux-ha.org/
[Linux-HA] Resource fail and node fence
Hi all, I've got a two node pacemaker/corosync cluster with some virtual domain resources on some DRBD devices. Every DRBD device is configured in dual primary setup and I have enabled the live migration. Cluster has also stonith enabled. My problem is that if a live migration for a single virtualdomain resource fails, then this node gets fenced, making unavailable also all the other virtual machines (that gets restarted on the other node after a poweroff). As I saw the way to make a single resource fail not fencing the node where it fails is to declare an on-fail=restart option for the virtual domain. Is it the correct approach or is there a more elegant way to obtain what I want? Thanks to all, -- RaSca Mia Mamma Usa Linux: Niente è impossibile da capire, se lo spieghi bene! ra...@miamammausalinux.org http://www.miamammausalinux.org ___ 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] What's wrong in my configuration for GFS2 under Pacemaker ?
Hi, only for information : It works fine now, it was just due to the fact that on chili1 , gfs2-utils rpm was not installed, and it is strange that I succeeded to mount the FS gfs2 without this rpm installed which install /sbin/mount.gfs2 ... And perhaps just a dependancy is missing in rpm gfs-pcmk-3.0.12-23.el6.Bull.1.x86_64 on gfs2-utils. Regards Alain De :alain.mou...@bull.net A : linux-ha@lists.linux-ha.org Date : 13/09/2011 08:59 Objet : [Linux-HA] What's wrong in my configuration for GFS2 under Pacemaker ? Envoyé par :linux-ha-boun...@lists.linux-ha.org Hi, ( Pacemaker 1.1.2.7 on RHEL6 ) I just tried to configure a GFS2 Filesystem under Pacemaker, but the FS is always mounted only on one side, not both sides : Clone Set: dlm-clone [dlm] Started: [ chili0 chili1 ] Clone Set: gfs-clone [gfs-control] Started: [ chili0 chili1 ] Clone Set: fsGFS2 [fs-GFS2] Started: [ chili1 ] Stopped: [ fs-GFS2:0 ] and after a while the node chili0 where the FS mount fails is fenced by chili1 Is there something missing or wrong in my configuration below ? or is there an already known issue on this subject ? Thanks Alain Moullé My configuration around gfs2 items: node chili0 node chili1 primitive dlm ocf:pacemaker:controld \ op start interval=0 timeout=90s \ op stop interval=0 timeout=100 \ op monitor interval=120s primitive fs-GFS2 ocf:heartbeat:Filesystem \ params device=/dev/sdl directory=/mnt/gfs2 fstype=gfs2 \ op start interval=0 timeout=60 \ op stop interval=0 timeout=60 \ op monitor interval=20 timeout=40 \ primitive gfs-control ocf:pacemaker:controld \ params daemon=gfs_controld.pcmk args=-g 0 \ op monitor interval=120s \ op start interval=0 timeout=90s \ op stop interval=0 timeout=120s clone dlm-clone dlm \ meta interleave=true clone fsGFS2 fs-GFS2 clone gfs-clone gfs-control \ meta interleave=true colocation gfs-with-dlm inf: gfs-clone dlm-clone order start-gfs-after-dlm inf: dlm-clone gfs-clone ___ 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] How to configure only logs warning or critical
Hi I wonder if someone could list where I can tune the log level info for all softwares involved when using the stack Pacemaker/corosync . I set debug:off in corosync.conf but there are still lots of logs information in syslog , generated by lrmd etc. Is it possible to have the exhaustive list of software levels where we can tune the log level ? Thanks Regards Alain ___ 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 and squid
Hi all, I've got a question for heartbeat. How can I made this : If squid stop or be killed on node1, how make node2 be master ? Actually, node2 become master only when node1 is down, or heartbeat service on node1 is down, but if I kill squid, nothing happen. I'm using Centos 6 and last heartbeat version. Thanks a lot for your responses ! -- Nicolas ___ 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 and squid
Hi, On Thu, Sep 01, 2011 at 06:30:46PM +0200, Nicolas Repentin wrote: Hi all, I've got a question for heartbeat. How can I made this : If squid stop or be killed on node1, how make node2 be master ? Actually, node2 become master only when node1 is down, or heartbeat service on node1 is down, but if I kill squid, nothing happen. I'm using Centos 6 and last heartbeat version. Using just heartbeat and no pacemaker? Only pacemaker has service monitoring. Thanks, Dejan Thanks a lot for your responses ! -- Nicolas ___ 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] only one node complains heartbeat over /dev/ttyS0
anyone has thoughts on it? Thanks. Hai Tao From: taoh...@hotmail.com To: linux-ha@lists.linux-ha.org Date: Mon, 12 Sep 2011 15:49:33 -0700 Subject: [Linux-HA] only one node complains heartbeat over /dev/ttyS0 I have a two node HA, and heartbeat is configured using a network LAN and a serial cable. However, I keep seeing this message in 02: WARN: glib: TTY write timeout on [/dev/ttyS0] (no connection or bad cable? [see documentation]) But on 01, everthing is ok: info: These are nothing to worry about. I have checked that both nodes have heartbeat running. My question is if 01 is receiving heartbeat from 02, and this is over a serial link, why 02 cannot receive heartbeat from 01 over /dev/ttyS0? Thanks. Hai Tao ___ 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] howto use heartbeat API
Hi,all I am program with heartbeat3.0.4, I have install it for three node, now I want to use heartbeat local IPC and cluster IPC to communicate with other node . And my question is how to compile, it always say that can't find hb_api.h ...hb_api_core.h.. how to wirte the makefile ? And how tu use local IPC and cluster IPC api to communicate ? one node want send message to other node , i have to code a CLI and a Daemon, daemon get Command from CLI from local IPC, and send message (the command from CLI) to other node via cluster IPC, HOW to do it ___ 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] howto use heartbeat API
On Thu, Sep 15, 2011 at 01:47:33AM +0800, 4070421 wrote: Hi,all I am program with heartbeat3.0.4, I have install it for three node, now I want to use heartbeat local IPC and cluster IPC to communicate with other node . And my question is how to compile, it always say that can't find hb_api.h ...hb_api_core.h.. how to wirte the makefile ? And how tu use local IPC and cluster IPC api to communicate ? one node want send message to other node , i have to code a CLI and a Daemon, daemon get Command from CLI from local IPC, and send message (the command from CLI) to other node via cluster IPC, HOW to do it Since the project is no longer one monolithic source repo. you need the cluster-glue sources/libraries/headers as well. (http://hg.linux-ha.org/glue) In the heartbeat source tree, there are lib/hbclient/api_test.c membership/ccm/ccm_testclient.c which may help you get going. I'm curious, what is your project about? If you stay with heartbeat, I strongly recommend to use latest mercurial (which is one changeset beyond 3.0.5). http://hg.linux-ha.org/dev/shortlog 3.0.4 to 3.0.5 contains only a handful of changesets. But they fix very bad behaviour in case you should ever have to deal with packet loss, for whatever reason. Have you considered using corosync instead? I'm not trying to push you either way, I'd just like to know what alternatives have been considered, and what you based your decision for heartbeat on. Cheers, -- : Lars Ellenberg : LINBIT | Your Way to High Availability : DRBD/HA support and consulting http://www.linbit.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
[Linux-HA] Strange behavior ldirectord direct routing
Hope someone here can give me some pointers. I've set up a ldirectord LinuxHA cluster. When I start it up its a simple set up. I have one VIP pointing to 2 backend servers and everything seems to come up ok. Output of ip addr: (weird set up I know - not my network tho just trying to get this working) 1: lo: LOOPBACK,UP,LOWER_UP mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo 2: eth0: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 5c:f3:fc:78:c9:98 brd ff:ff:ff:ff:ff:ff 3: eth1: BROADCAST,MULTICAST mtu 1500 qdisc noop qlen 1000 link/ether 5c:f3:fc:78:c9:9a brd ff:ff:ff:ff:ff:ff 4: usb0: BROADCAST,MULTICAST mtu 1500 qdisc noop qlen 1000 link/ether 5e:f3:fc:7b:c9:9b brd ff:ff:ff:ff:ff:ff 5: eth0.101@eth0: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc noqueue link/ether 5c:f3:fc:78:c9:98 brd ff:ff:ff:ff:ff:ff inet 172.28.89.46/25 brd 172.28.89.127 scope global eth0.101 inet 172.28.191.155/25 brd 172.28.191.127 scope global eth0.101 --- This is the VIP The VIP above points to the backend servers 172.28.191.139 and 172.28.191.140. Without HA running I can telnet to the backend servers to the proper port (8080 in this case). As soon as I start up HA and ldirector, I am unable to telnet to the backend servers on port 8080 on the node that is primary. I can telnet to them on the backup node and on any other server on the network but not from the primary node. I don't have access to the backend servers but the SA tells me he set up the loopback with an alias to the VIP as I asked and he also said he added the appropriate entries to prevent the arp issues into sysctl.conf on the backend servers. So what the heck is going on? Why does HA suddenly make it impossible to connect to the backend servers from the primary node? Another problem I have is that I cannot telnet to the VIP on port 8080 no matter what node it is running on. I think if I can resolve the problem above, this one will go away too. Thanks -mike ___ 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] Strange behavior ldirectord direct routing
Replying to my own post here but the fact that the VIP is on a different subnet has me curious if this could be an issue: ip addr show eth0.101 5: eth0.101@eth0: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc noqueue link/ether 5c:f3:fc:78:c9:98 brd ff:ff:ff:ff:ff:ff inet 172.28.89.46/25 brd 172.28.89.127 scope global eth0.101 inet 172.28.191.155/25 brd 172.28.191.127 scope global eth0.101 -- VIP On 11-09-14 06:37 PM, mike wrote: Hope someone here can give me some pointers. I've set up a ldirectord LinuxHA cluster. When I start it up its a simple set up. I have one VIP pointing to 2 backend servers and everything seems to come up ok. Output of ip addr: (weird set up I know - not my network tho just trying to get this working) 1: lo:LOOPBACK,UP,LOWER_UP mtu 16436 qdisc noqueue link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo 2: eth0:BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 5c:f3:fc:78:c9:98 brd ff:ff:ff:ff:ff:ff 3: eth1:BROADCAST,MULTICAST mtu 1500 qdisc noop qlen 1000 link/ether 5c:f3:fc:78:c9:9a brd ff:ff:ff:ff:ff:ff 4: usb0:BROADCAST,MULTICAST mtu 1500 qdisc noop qlen 1000 link/ether 5e:f3:fc:7b:c9:9b brd ff:ff:ff:ff:ff:ff 5: eth0.101@eth0:BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc noqueue link/ether 5c:f3:fc:78:c9:98 brd ff:ff:ff:ff:ff:ff inet 172.28.89.46/25 brd 172.28.89.127 scope global eth0.101 inet 172.28.191.155/25 brd 172.28.191.127 scope global eth0.101 --- This is the VIP The VIP above points to the backend servers 172.28.191.139 and 172.28.191.140. Without HA running I can telnet to the backend servers to the proper port (8080 in this case). As soon as I start up HA and ldirector, I am unable to telnet to the backend servers on port 8080 on the node that is primary. I can telnet to them on the backup node and on any other server on the network but not from the primary node. I don't have access to the backend servers but the SA tells me he set up the loopback with an alias to the VIP as I asked and he also said he added the appropriate entries to prevent the arp issues into sysctl.conf on the backend servers. So what the heck is going on? Why does HA suddenly make it impossible to connect to the backend servers from the primary node? Another problem I have is that I cannot telnet to the VIP on port 8080 no matter what node it is running on. I think if I can resolve the problem above, this one will go away too. Thanks -mike ___ 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
