Tuesday, January 13, 2009

Failure at final check of Oracle CRS stack. 10

Who Someone run "root.sh" on second node and get this error:

> Failure at final check of Oracle CRS stack. 10

that is the problem about Firewall or some securities.

With Linux/Unix OS, that should check IPTABLES or Packet Filter.

and make sure on every nodes of Cluster , that can communicate on Public IP (TCP) and Private IP (UDP) with every nodes.

> iptables -L

8 comments:

Anonymous said...

Hey,
On redhat 5 i still get Failure at final check of Oracle CRS stack. 10
on node 2,
even though iptables -L outputs
Chain INPUT (policy ACCEPT)
target prot opt source destination

Chain FORWARD (policy ACCEPT)
target prot opt source destination

Chain OUTPUT (policy ACCEPT)
target prot opt source destination

I belive everyhting should be ok, network wise, I can ping both nodes on all IPs (except my VIP).

Any ideas where to look for sollution on node2?

-Andreas

Surachart Opun said...
This comment has been removed by the author.
Surachart Opun said...

You should check Interconnect (UDP on Linux)

between run root.sh script... check:

Example:
$ ps -ef |grep crs
...
/bin/su -l oracle -c /usr/app/oracle/product/10.2.0/db_1/bin/crsctl check boot > /tmp/crsctl.3960
$ ps -ef | grep cssd
/bn/sh /etc/init.d/init.cssd startcheck

After that, Check /tmp/crsctl.3960 file.

Or Check OCR by ocrcheck command-line.

If you need to recreate (239998.1), can help

Anonymous said...

Thanks!

It turned out that my network card in ESX was enabled, but not powered on. So in Linux everything looked fine.
Thanks for pointing me in the correct direction :)

-Andreas

Surachart Opun said...

Failure at final check of Oracle CRS stack. 10

If it's not the problem with Interconnect Network.

We may use "ocrcheck" to check id both nodes
Example:

oracle@rac1 bin$ ./ocrcheck
Status of Oracle Cluster Registry is as follows :
Version : 2
Total space (kbytes) : 10482292
Used space (kbytes) : 304
Available space (kbytes) : 10481988
ID : 749094372
Device/File Name : /dev/raw/raw1
Device/File integrity check succeeded

Device/File not configured

Cluster registry integrity check succeeded

oracle@rac2 bin$ ./ocrcheck
Status of Oracle Cluster Registry is as follows :
Version : 2
Total space (kbytes) : 10482292
Used space (kbytes) : 304
Available space (kbytes) : 10481988
ID : 1155650855
Device/File Name : /dev/raw/raw1
Device/File integrity check succeeded

Device/File not configured

From "ocrcheck"
rac1 -> ID : 749094372
rac2 -> ID :1155650855

So, check about binding... raw device

If you use multi-path ... you should bind the raw devices to the multipath devices name.
refer:metalink 344994.1

Anonymous said...

Can you elaborate how to check UDP on linux?

I am still getting the issue after running root.sh on the second node

thanks

Edgar

Surachart Opun said...

on linux "tcpdump" can dump packet ...

If you sure you no policy firewall, bu still error... you should another case

# iptables -F
# iptables -L

- check Check /tmp/crsctl.* while run "root.sh" command

- run "ocrcheck" to check ID (make sure you use OCR/VOTE on the same disk both nodes)

Anonymous said...

Thank you for saving my day! This works after flushing...