Question about cluster and HA-Oracle maintenace
Hi Cluster guru's,
I am new to sun cluster and HA-ORACLE too, If someone answer the following question with the command and procedure, it will be a greate help.
I have one HA-Oracel RG with with all required RS, I need to do following things .....
1. DBA wants to do some maintenace on database so they need to start/stop database several time, so i want to give them database without cluster funtionality, so that when they close the database it should not move to other node.
2. By mistake one DBA stop the database and it strated on other node, now i want my database back to original node when i gave scswitch -z -g <RG-Name > -h <Node_Name>
i got the error something like this...
This RG already failed on this node and may be faild on other node too, cannot start the RG
3. On both the node I can not start the RG, there is no primay node for RG, error mesg like above
4.Just want to mount the file system for maintenace like fsck etc..., do not want tostart the database ?
thanks & regards
[1074 byte] By [
klk] at [2007-11-26 8:36:57]
# 1
Hi,
some suggestions:
1) you can disable the DB monitoring using scswitch -n -m -j ha-oracle-resource
2,3 ) check console or /var/adm/messages on both nodes. There should be some info about the error.
4) Don't understand clearly your question, however I'll try with some info.
Usually FS is being managed from SUNW.HAStorage+ resource. If it's a global FS, it's mounted at boot, if it's a FFS (Failover FS) is mounted from HAStorage+ resource.
If you don't want to start the database simply issue: scswitch -n -j ha-oracle-resource
However the FS will be already mounted. you can use scswitch -n -j HAStorage+-resource to disable it and manage your FS manually. Again check what kind of FS (global or FFS) you have.
Hope this help
regards
gs
# 2
thanks for your time, in my question #2&3 what I mean is
there were so many occasions, what happen i stop the RG and it moved to other node then i tried again to move on the same node it fails there and state that it is already failed on this node, after that when i reboot that node where orignally RG was working, it start the RG.
or
lets think dba does not know about the cluster and he stop the DB on first primary node and when he saw DB moved on other node it stop the DB there also, in this case you can not start the db in any of the node, I faced one time like this, so i put the RG in umanged mode then again in managed mode after that i was able to open RG, but i dont know exactly the procedure it was just hit-and-trail.
klk at 2007-7-6 22:08:35 >
