The number of messages in Export queue not going down after a config was activated with wrong tar (4299430)

The number of messages in Export queue not going down after a config was activated with wrong tar

A config file was activated using the wrong hostname for target. This resulted in "show" output from sp_ctrl showing two Export processes whereas the qstatus showed only one Export queue. When the customer realized the mistake, he/she created a config file with correct target hostname and activated the new config file which successfully activated. The # of messages in Export queue keep increasing in one direction though the messages are posting fine to the target. Normally the messages should be Readrelease from Export queue, once they have made it to target but in this case there are two readers for the Export queue (two sp_xport processes, one going to the correct host whereas another going to unknown host). Unless the redundant reader is removed, the messages will keep growing in Export though they will also post. The event log shows the following type of messages over and over:

10/12/06 12:47 Process launched: sp_xport (to sfprdbs002_adm queue sjprdbs001_adm) [pid = 30604]
10/12/06 12:47 Cannot connect to host sfprdbs002_adm Connection refused

A config file was activated using the wrong hostname for target.

Use the qview deluser command to remove the extra reader for the wrong route (in our case the extra Export process) after shutting down Shareplex. The solution # 17611 provides details on it. The steps can be outlined as follows:

./qview -i
qview> qsetup
qview> list (this will give you the queue names and their respective readers and will help you in identifying the redundant reader to be removed)
qview> deluser <export queue name> <username>
qview> exit
Then restart Shareplex