Log wrap detected on non-ASM (4269528)

1. Archive log was removed or none detected
2. NOARCHIVELOG mode, therefore on-line redo logs maybe overwritten before capture was able to parse the logs.

By looking into the source $SP_SYS_VARDIR/log/event_log for possible indication of the problem. If there are errors similar to "log wrap detected" and/or "redo log could not be parsed", then you have a log wrap situation. These errors are usually accompanied by a log sequence number xxx.

1. Make sure the database is in archivelog  mode.

2. Find the missing sequence log number xxx and put it back in its original directory.

Note: If the logs are still on a mountable partition, there is a native Shareplex parameter that will allow you to point to multiple location.

i.e: Set param SP_OCT_ARCH_LOC  /disk1/log

If RAC, you will need to add the path where each node has its archivelogs:

sp_ctrl> set param SP_OCT_ARCH_LOC  /disk1/path/instance1;disk2/path/instance2;/disk3/path/instance3

Take effect: Immediately

3.  If the missing logs are unrecoverable, you have two options

a. stop capture and update shareplex_actid on source and advance the seqno to the next available seqno and compare/repair the out of sync tables

update shareplex_actid set seqno=xxx, offset=0, log_start_offset=0;

commit;

if you are on RAC, you have to update the corresponding entries for the related RAC node.

b. Cleanup replication and reinitialize target by running $SP_SYS_PRODIR/bin/ora_cleansp, reactivate the config_file and resync the Out of Sync tables by taking hot backup and reconcile the appropriate sequence or scn number.