Every cluster is different .. I've seen some database clusters that the only thing the cluster controls is the filesystems. (Like that's not a disaster waiting to happen.. ). In that particular case, off-lining cluster resources without DBA involvement could make for a bad day Since it looks like you might not be familiar with the nuances of this cluster, here's what I consider the safe route for DB servers:
Do hastatus -sum, and note the group that controls the database. Then look at
Code:
/etc/VRTSvcs/conf/config/main.cf
and see what that group actually does (or look via the hagrp and hares commands). Assuming the database itself is controlled by the cluster, bring it down like this -
In one window:
Code:
tail /var/VRTSvcs/log/engine_A.log
In another:
Code:
hastop -g <database group> -sys <system its running on>
Watch hastatus, and/or the log file you're tailing. If things go down smoothly, then great. If it hangs up waiting on the DB, let the DBA do their thing. The log will usually tell you everything you need to know. Be patient, depending on the DB, it can take a long time to come down.
Once the cluster and the DBA are both satisfied that the DB is down, you can usually then do a hastop -all, and the cluster should pretty easily take care of the dependencies. Wait for it to complete, and help it along if necessary using the info from the log file you're tailing.
Personally, if I'm not 100% comfortable with the system I'm on, I'm paranoid. In that case I like to do everything with cluster nodes one at a time. So offline all resources on all nodes of the cluster and stop VCS, then shut down one node, then the next, etc. Same on the way up. Bring up nodes one at a time, if you want to be extra careful. Let VCS find its brain on one node before another tries.
I've brought down CFS nodes at the same time, and end up with goofy fencing issues. (in hindsight I should have fully closed out gab and llt). It's never happened when I bring them down one at a time, so if I have the time, I like to do it that way.
Anyway, hastop -all would probably work just fine on a properly configured cluster. The fun is when the cluster isn't properly configured. And unless you know for sure either way, it's best to play it safe.
Last edited by cubemonkey; 12-10-2011 at 12:55 AM..
Hi ! I have a simple setup of 2 PC (with linux Red-Hat) where the first PC is the primary machine and the second the backup. I use DRBD for data replication and Red-Hat cluster suite for HA (High Availability). I have tested both.
Now I NEED a COMMON IP ADDRESS (or Master/unique IP address) for... (0 Replies)
Hi ! I have a simple setup of 2 PC (with linux Red-Hat) where the first PC is the primary machine and the second the backup. I use DRBD for data replication and Red-Hat cluster suite for HA (High Availability). I have tested both.
Now I NEED a COMMON IP ADDRESS (or Master/unique IP address) for... (3 Replies)
hello Gurus,
My current set up is 3 to 1 Cluster (SUN Cluster 3.2) running oracle database. Task is to reboot the servers. My query is about the procedure to do the same.
My understanding is suspend the databases to avoid switchover. Then execute the command scshutdown to down the cluster... (4 Replies)
Hi all, i have 3 nodes cluster (Centos 5 cluster suit) with out quorum disk,
node vote = 1,
the value of a quorum = 2,
when 2 nodes going offline, cluster services are destoys.
How i can save the cluster and all services(move all services to one alive node)
with out quorum disk when other... (3 Replies)
I am new to setting up sun solaris 10 cluster, I have 2 sun sparc t3-1 servers (identical), going to use them as web servers (sun one java web server 7), looking for data replication and real time fail over. My question is do I need external storage to configure the cluster? or I can just use... (3 Replies)
Hi all.
May I get some expert advice on troubleshooting performance issues of a 1000 nodes Apache LB cluster. Users report slow loading/response of webpages. Different websites are hosted on this cluster for different clients. But all are reporting the same issue.
Could you please let me know... (1 Reply)
Hi,
A customer I'm supporting once upon a time broke their 2 cluster node database servers so they could use the 2nd standby node for something else. Now sometime later they want to bring the 2nd node back into the cluster for resilance. Problem is there are now 3 VG's that have been set-up... (1 Reply)
Is there any way to create a arbitrary node for ocfs2 on a virtual machine (others are physical servers) so it won't go panic when one of physical server goes down?
This is for load balanced application servers.
Any setting example or tips?
Thanks. (0 Replies)
Hi All;
I try to build a Redhat Cluster (CentOS 6) on vmware. But each node sees the other down like:
# clustat
Cluster Status for mycluster @ Wed Apr 8 11:01:38 2015
Member Status: Quorate
Member Name ID Status
------ ---- ... (0 Replies)
Hi All;
I try to build a Redhat Cluster (CentOS 6) on vmware. But each node sees the other down like:
# clustat
Cluster Status for mycluster @ Wed Apr 8 11:01:38 2015
Member Status: Quorate
Member Name ID Status
------ ---- ... (1 Reply)
Discussion started by: Meacham12
1 Replies
LEARN ABOUT DEBIAN
votequorum_leaving
VOTEQUORUM_LEAVING(3) Corosync Cluster Engine Programmer's Manual VOTEQUORUM_LEAVING(3)NAME
votequorum_leaving - Tell other nodes that we are leaving the cluster
SYNOPSIS
#include <corosync/votequorum.h>
int votequorum_leaving(votequorum_handle_t handle);
DESCRIPTION
The votequorum_leaving function is used to tell the other nodes in the cluster that this node is leaving. They will (when the node actually
leaves) reduce quorum to keep the cluster running without this node.
This function should only be called if it is known that the node is being shut down for a known reason and could be out of the cluster for
an extended period of time.
Normal behaviour is for the cluster to reduce the total number of votes, but NOT expected_votes when a node leave the cluster, so the clus-
ter could become inquorate. This is correct behaviour and is ther eto prevent split-brain.
Do NOT call this function unless you know what you are doing.
RETURN VALUE
This call returns the CS_OK value if successful, otherwise an error is returned.
ERRORS
The errors are undocumented.
SEE ALSO votequorum_overview(8), votequorum_initialize(3), votequorum_finalize(3), votequorum_dispatch(3), votequorum_fd_get(3),
corosync Man Page 2009-01-26 VOTEQUORUM_LEAVING(3)
12-10-2011
Since it looks like you might not be familiar with the nuances of this cluster, here's what I consider the safe route for DB servers: