Hot answers tagged

17

Ok seeing as how there's nothing but tumbleweeds gusting through this thread I went out a did the legwork myself. I've made the decision to go with Corosync for the following reasons: Corosync is a newer project (based on openais) and seems to be the 'future' for messaging layers Hearbeat looks like it will be maintained for quite a while longer but ...


14

This is a slightly older question but the problem presented here is based on a misconception on how and when failover in clusters, especially two-node clusters, works. The gist is: You can not do failover testing by disabling communication between the two nodes. Doing so will result in exactly what you are seeing, a split-brain scenario with additional, ...


12

I am actually working on something very similar to what you described (a fail-over cluster on EC2), and after trying out Heartbeat, settled on Corosync as my messaging layer. Corosync will run on multiple servers and it does support Unicast (UDPU) as of version 1.3.0 (from Nov, 2010). I have setup and tested Corosync on Amazon's EC2 cloud (using Amazon's ...


12

I prefer to use keepalived for high-availability. I find it simpler to setup (one daemon and config) than heartbeat and company. The only drawback I run into, is that keepalived doesn't have a unicast option by default, and only uses VRRP for communication (The author of HAProxy has written a unicast patch for keepalived however)


9

Sorry, but the part about Pacemaker is not true. The Pacemaker regression and release tests make extensive use of automation. To configure without an active cluster, prefix all commands with CIB_file=/var/lib/heartbeat/crm/cib.xml or set it in your environment. Just be sure you remove the .sig file before starting the cluster. For clusters without quorum, ...


7

Red Hat has switched to use pcs (Pacemaker Configuration System): Name : pcs Arch : noarch Version : 0.9.26 Release : 10.el6 Size : 248 k Repo : installed From repo : base Summary : Pacemaker Configuration System URL : http://github.com/feist/pcs License : GPLv2 Description : pcs is a corosync and ...


5

You can use RFC1918 IP addresses to test your configuration. Depending on the load balancing setup, the use of RFC1918 addresses may continue in some form into production.


5

It turns out the solution is quite simple all that is needed is to set 'start-failure-is-fatal' to false. crm configure property start-failure-is-fatal=false This is a global option and it's default is true hence my failing resource. Pacemaker configuration reference


5

Is it possible to use Pacemaker in top of Hyper-V for now? Or may be exist another way to use DRBD in dual primary? It's possible but resulting performance is not going to be great TBH. Get some native Windows solution like StarWind Virtual SAN you can run right on top of Hyper-V. These particular guys are free for two nodes, you may need to ask for "hyper ...


4

You can use the cibadmin utility to query the Pacemaker configuration, move the primitives around accordingly within the <group id="group-vol1"></group> tags, and then push the new config back into Pacemaker. The process would look something like this: # cibadmin --query > /tmp/cib.xml # vi /tmp/cib.xml # cibadmin --replace --xml-file /tmp/...


3

NFS stores a lot of the client's state on the server. Pacemaker/OpenAIS can't make up for NFS's shortcomings in this area. The grace period is there for the server and clients to recapture state. It's part of the protocol. Anyway, it seems that you are not moving client state over completely (like /var/lib/nfs contents). See this for ideas and what needs to ...


3

In the open source world, there's RedHat Cluster Suite. It's been several years since I've implemented RHCS so I don't have many relevant things to say about it today. Commercially, there is Veritas Cluster Server. No experience with it. A much simpler and open source HA tool is UCARP. UCARP doesn't provide nearly the same kind of "infrastructure" that ...


3

My gut level instinct is to say no, those are really not the right tools for cluster management on EC2. I've used them on stand alone hardware and found you have to have a very specific set of needs / failure cases for them to really make sense there. I cannot concoct a use case in my head that would demand those tools over more specific cloud monitoring ...


3

I'd go with GlusterFS. The latest version 3.x supports geo-replication (long latent pipe type of thing) as well as LAN replication. There's plenty of docs about how to replicate and spread data across the cluster. I don't like DRDB, because there's a limit on the number of nodes you can use. I think GlusterFS on decent hardware, with a decent bit of ...


3

Your cluster architecture confuses me, as it seems you are running services that should be cluster-managed (like Varnish) standalone on two nodes at the same time and let the cluster resource manager (CRM) just juggle IP addresses around. What is it you want to achieve with your cluster setup? Fault tolerance? Load balancing? Both? Mind you, I am talking ...


3

I am late to reply to your question, but here goes: Yes heartbeat is deprecated. No this is not a distro specific problem There are a lot of benefits to using Corosync instead of heartbeat, the first and foremost is point number 1 above. Still, I will list as many as I know of here. Comparison of features: First off the only benefit(IMO) of using ...


3

In the sources of pcs-0.9.115-32.el7 the getClusterName function is implemented: def getClusterName(): if is_rhel6(): try: dom = parse(settings.cluster_conf_file) except (IOError,xml.parsers.expat.ExpatError): return "" return dom.documentElement.getAttribute("name") else: try: f = ...


3

Definitely. Create a configuration file (named 'cib.txt', in our example) with the same syntax you've used in your example commands: primitive VIP ocf:heartbeat:IPaddr2 params ip=10.0.2.200 nic=eth0 \ op monitor interval=10s timeout=20s \ op start interval=0 timeout=20s \ op stop interval=0 timeout=20s Then you can load that file using the ...


3

res_exportfs_export2_stop_0 on xx.xx.xx.1 'unknown error' (1): call=47, status=Timed Out, last-rc-change='Tue Mar 31 12:53:04 2015', queued=0ms, exec=20003ms Shows that your res_exportfs2 resources failed to stop due to a timeout. It may simply be that it needs a longer timeout. Try configuring a stop timeout for this resource like so: primitive ...


3

I don't believe this is a question of instructing pacemaker to only use the virtual IP, but rather telling Nginx to only use the virtual IP. I am no Nginx expert, but reading through their documentation I found a configuration directive named proxy_bind. I believe this might be what you need to set to get the desired behaviour. From the Nginx documentation ...


2

I am currently testing "stretch cluster" using Red Hat Cluster Suite and DRBD. I am typing this at a hotel near Red Hat Summit in Boston which just ended. I talked with the Red Hat CLuster Suite developers and they said stretch clusters were not supported at this time. This won't stop me from working on it for fun though. My set up is four HP blades in a ...


2

DRBD is dead slow as everybody knows. You can't use that for high load enterprise purposes. It uses 128 KiB hashing functions which limit the IO requests to max. 128 KiB instead of 512 KiB what a regular HDD can provide. Furthermore, there is a stupid IO request size detection. This thing only works when connected to the other host. If you loose the ...


2

Whoever in curious to know if heartbeat with more than 2 node is possible, then I say heartbeat is not limited to 2 nodes we can join more than 2 nodes. Actually I didn't copy the same authkey for the 3rd node. All the three nodes is listing now. I Also tried Corosync, and one of the node I couldn't get online don't know why having all config and keys same, ...


2

Well, this is what happens when you don't read the manual carefully. The colo constraint colocation xen-db -inf: xen-db01 xen-db02 means that xen-db01 will be placed relative to xen-db02. So really I should have been starting xen-db02, and THEN xen-db01. Doing crm resource start xen-db02 crm resource start xen-db01 starts the resources normally, ...


2

I haven't seen how 6.1 changes things, I've only done it with 6.0, but this is what worked for me: Install the following patterns: Scalable File Systems Resilient Storage High Availability On each node, start the ricci service (service start ricci), entering a password when prompted Start the luci service (service start luci) on one node Go to port :8084 ...


2

First of all, let me say that I've set up quite a few clusters over the last decade, and I've never seen one where there was a dependency like you've described. Usually one would set it up so that the services provided don't depend on which host is active and which is standby and you don't care which host has the resource, as long as it's up on one of them. ...


2

I think problem is not in cluster configuration, but in your routing architecture. VIPArip resource agent manages local quagga to send routing updates. But you also need to use this routing updates to change routes to point to active server. I'll try to explain how it works. Look at the picture. HA1 and HA2 is linux-ha cluster members with quagga running....


2

Eureka! Both of us forgot a very very important log file, it's... /var/log/mysqld.log: socket: '/var/lib/mysql/mysql.sock' port: 3306 MySQL Community Server (GPL) by Atomicorp [Note] Slave SQL thread initialized, starting replication in log 'mysql-bin.000082' at position 58569, relay log './mysqld-relay-bin.000002' position: 58715 [Note] Slave I/O thread:...


2

The spammy log messages you see are coming from the ocf:heartbeat:exportfs resource agent. They appear every 30 seconds, which corresponds to the monitoring interval you specified in the exportfs primitive definitions. The resource agent is a bit too verbose, IMHO, but this should not be a problem. Just make sure you logrotate often enough that the logs don'...



Only top voted, non community-wiki answers of a minimum length are eligible