| Logical Standby Archive log rejected by RFS client [message #307718] |
Wed, 19 March 2008 12:06  |
ylatey
Messages: 1
Registered: March 2008
Location: India
|
Junior Member |
|
|
Hi!
I have configured single instance logical standby from two node RAC. It was working fine, but in between it started giving this message: "Archive log rejected (thread 2 sequence 2816) by RFS clients" also similar message for two other log sequences.
These archives are present in standby_archive_dest area. However many more files are archived to this location after these and standby is working fine, but i am continuously getting this message for 3 archives every 45 mins in alert log on standby.
I tried to register these files as: alter database register logical logfile..." but this throws error saying file already registered.
Anybody faced this issue or have any idea how to resolve this, please help at earliest.... Thank you very much in advance
Yogesh
|
|
|
|
| Re: Logical Standby Archive log rejected by RFS client [message #309147 is a reply to message #307718] |
Wed, 26 March 2008 10:56  |
RichB
Messages: 3
Registered: March 2008
Location: California
|
Junior Member |
|
|
I've resolved this multiple ways... First of all, the RFS (Remote File Server) appears to get the log just fine, but somehow it is not completely intact.
At least Oracle keeps shipping the others so when you get going again you'll have a flury of archives to apply.
It can related to a network interruption.
We normally do the following:
Perform a checksum of both archive logs (the one on the primary and the standby. If not equal, then you need to reship.
You can also run a BACKUP VALIDATE ARCHIVELOG SEQUENCE <seq#>; on the standby to verify it's integrity.
Reship your archive log in question to another directory on the standby. Once there, do a manual registration of the file.
ALTER DATABASE REGISTER OR REPLACE LOGFILE'<fqn of archlog>;
It it doesn't work, then we do this with the standby out of recovery ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL; , cp the file again, and register manually, then resume recovery with either
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT; or ALTER DATABASE RECOVER MANAGED STANDBY DATABASE THROUGH SEQUENCE <#> NODELAY; to get through a specific log.
Also, do you have the 'REOPEN command in your log_archive_dest? We set ours to 300 seconds, to reopen the connection.
|
|
|
|
] 
Blog
Search
Help
Members
Register
Login
Home
