tnsping Ok but sqlplus hangs

one of the many reason

doing strace . it was showing as below in infinite loop


times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
times(NULL)                             = -2093807697
               = -2093807696
[oracle@b6jasmine-prbs ~]$

the server was up for 200 days . so i had to do a reboot after that it worked fine ...

goldengate vip getting connection refused

goldengate bidirectional replication on site A,site B,site C

the goldengate vip is configured on site A .  but this morning when i came the vip was running on node B  on site B

and  the pump process running on site B and site C  that send to site A  were failing with

WARNING OGG-01223  Oracle GoldenGate Capture for Oracle, dpsiteA.prm:  TCP/IP error 111 (Connection refused), endpoint: :7809.

and trying a telnet connect from site B to site A    telnet 7809  was returning

telnet 7809

Trying ...

telnet: connect to address : Connection refused

telnet: Unable to connect to remote host: Connection refused

so on relocating the vip from node B to node A on site A   the issue was resolved.  

Goldengate replication on RAC and timezone change

I was troubleshooting a goldengate issue in a 2node RAC on RHEL8   i ran into a issue
where extract was not writing to trail log .

We recently changed the timezone of the server to GMT  from PST to match with other target server residing on different continent  while setting up the environment.   and we did our regular approach of bringging down the cluster and database using crsctl/srvctl way.  however i was surprised to see that the redo log timestamp was still on PST . Which means the golden gate extract begin now was not finding any latest change from the redo log threads .   no matter how many times I brought up brought down the database using srvctl didnt change the online redo log timestamp.  so i did  a manual restart of database instances  after that the redo started showing the timestamp on GMT . and bouncing the manager resolved the extract issue .

ORA-609 on the alert log with TNS lost contact on the client

I ran into this issue on a  2 node RAC 11.2.0.3 cluster   while accessing from the client PC.  here the scan listener was using a port 1590 while node listener was running on 1521 .   checking with the network team regarding firewall port opening for 1590 was verified. The information from network team regarding port 1521 was not clear  while troubleshooing over the weekend.. First of all looking the ORA-609  on the alert log was making the issue looks very confusing . I was seeng the request coming to the server on scan listener log .   It took a lot of hours  with no success .  so thought of giving a try to convert the node listener port  to also using 1590 port. which  did the trick .