tnsping intermittent slowness and sometime throw TNS-12535
klnghauDec 9 2010 — edited Aug 24 2012Hi All,
We're encounter a very weird scenario in 3 nodes RAC environment. We encounter intermittent slowness in client environment.
sometime it will throw ORA-12170: TNS:Connect timeout error message when connect via SQLPLUS.
tnsping, sometime it fail, sometime very bad response (take more than 60000 msec response) and sometime it throw TNS-12535: TNS:operation timed out,
the result is not consistent.
Environment as below:
Oracle version: 11gR2 (11.2.0.1.0 - 64bit Production)
Os::Red Hat Enterprise Linux Server release 5.3
3 nodes RAC environment. (one node oralce not running at this moment)
This is 3 nodes RAC environment. We start getting this issue after one node went down.
The application become very slow, and keep hitting ORA-12170 error. We did not change anything, just node2 oracle is not running this moment.
I'm not sure whether this is Oracle Net service issue or network issue or DNS issue.
Assume SCAN-listener IP info as below:
nslookup scan-rac
Server: ipserver1
Address: ipserverx
Name: rac-scan.com
Address: ip1
Name: rac-scan.com
Address: ip2
Name: rac-scan.com
Address: ip3
We tried below:
1. ping rac-scan.com ==> we;re able to ping without any issue.
2. ping 3 individual ip ==> we're able to ping without any issue
3. tnsping rac-scan.com 5
==> It's intermittent, sometime all 5 OK, but sometime few taking long time, and sometime encounter TNS-12535 error. Not consistent, don't know the reaseon.
4.We have checked listener all running properly in node 1 and 3 (node 2 oracle not running)
5. When llogin via sqlplus, 2 out of 6, will throw error ORA-12170. We did not change any tnsname/sqlnet.ora file, all same before node 2 down.
And user complain application start hitting this error after node2 went down.
May I know anyone facing this kind of issue before? Please share with me, thanks in advance.
Regards,
klnghau