Export queue corruption and cannot read any message (21207)

Export queue corrupted

Process exited sp_xport (to crmtstrpt02.isus.emc.com queue [pid = 6860] - exit(1)
Error: sp_xport: failed to read export queue - exiting [sp_xport/6860]
Notice: Error reading queue Q+X, subqueue 0 [sp_xport(que)/6860]
Notice: peekahead failure buffer hdr=0xf18112b8 addr=0xf18551c8 [sp_xport(que)/6860]

This is mainly because of out of disk space in $SP_SYS_VARDIR/rim directory.

For version 5.1.6+, 5.3.x, try this work around first.

SOLUTION 1:

1. shutdown shareplex
2. qview -i
qview> fixup all
3. restart shareplex

For version 5.1.1 to 5.1.5, do not use fixup all due to some bugs. Instead use uscan to release corrupt messages.

SOLUTION 2:

1. shutdown shareplex
2. qview -i
qview >qsetup port#
qview >open x r
qview >uscan 0 1 3
qview >rrls 0 <qseq number returned from last command>
qview >exit

3. start sp_cop and export.

SOLUTION 3:

If uscan cannot fix the queue corruption, try reset export and post queue

On the Source machine

1. shutdown shareplex
qview -i
qview> qsetup
qview> reset x
qview >exit

On the Target machine

1. shutdown shareplex
2. qview -i
qview>qsetup port#
qview >reset p
qview >exit
If you have multiple post queues, reset the post queues that got corrupted.

3. sqlplus splex/splex
truncate shareplex_trans
4. restart shareplex

For 5.1.1 to 5.1.5, do not use fixup all. Instead use uscan.
For 5.1.6+ and 5.3.x, use fixup all first.