06-27-2012 01:11 PM
Hey Everyone,
I have a cluster main.cf file as follows:
group BATCH-SG (
SystemList = { foo = 0, bar = 1 }
AutoStartList = { foo }
OnlineRetryLimit = 3
)
IP ip3-RES (
Device = bond0
Address = "0.0.0.1"
NetMask = "255.255.254.0"
)
NIC nic1-RES (
Device = bond0
)
ProcessOnOnly DB_Watchdog-RES (
PathName = "/app/watchdog/bin/db-watchdog-wrapper"
UserName = ezbatch
)
ProcessOnOnly NAS_Watchdog-RES (
PathName = "/app/watchdog/bin/nas-watchdog-wrapper"
UserName = ezbatch
)
requires group Cluster_Storage-SG online local firm
ip3-RES requires nic1-RES
// resource dependency tree
//
// group BATCH-SG
// {
// ProcessOnOnly DB_Watchdog-RES
// ProcessOnOnly NAS_Watchdog-RES
// IP ip3-RES
// {
// NIC nic1-RES
// }
// }
I'm trying to set it up so that when either of the ProcessOnOnlys die, the group dies and VCS initiates a failover. However when either of the watchdogs exit with status 100, VCS just restarts them. Any ideas on what I'm doing wrong?
Solved! Go to Solution.
06-27-2012 01:21 PM
Ah, just seen you have OnlineRetryLimit set on group so you need to set this to zero:
haconf -makerw hagrp -modify BATCH-SG OnlineRetryLimit 0 haconf -dump -makero
Then VCS will failover group.
Mike
06-27-2012 01:18 PM
Check the value of RestartLimit:
hatype -display ProcessOnOnly -attribute RestartLimit
If this is non-zero then you need to set it to zero if you don't want VCS to restart and failover straightaway:
hatype -modify ProcessOnOnly RestartLimit 0
Mike
06-27-2012 01:21 PM
Ah, just seen you have OnlineRetryLimit set on group so you need to set this to zero:
haconf -makerw hagrp -modify BATCH-SG OnlineRetryLimit 0 haconf -dump -makero
Then VCS will failover group.
Mike
06-27-2012 02:00 PM
Thank you that fixed the issue.