Recover down chunks

[nAtHaNs]

hello all,

I have cloned our production server, system and database, to another server for testing (back-up and restore, disaster recovery).

But when I try to bring up the database, there are some chunks that are down. How do I recover them?

thank you.

 

[WiccaChic]

onspaces -s <spacename> -p <pathname> -o <offset> -O

for example:
onspaces -s dbs20 -p /dev/chunk112 -o 128 -O

 

[nAtHaNs]

thanks for replying WiccaChic.

How do I know the value of my offset?

 

[WiccaChic]

Thats a good question and I am afraid I do not know the answer. The only reason I know is because I have copies of the commands I have used as well as our former dba used when chunking out drives.

Sorry, hopefully one of the real gurus will respond to this.

 

[mynwo]

Hi,

Run Onmonitor, Select Dbspace & Info

 

[vpshriyan]

Hi,

Execute onstat -d to get a list of spaces and chunks and concentrate on the columns "offset" and "flags". The chunk/s which are non-operational will have tagged as "D"own in the flags and look at the offset column for the corresponding row in the chunk list.

Normally, for all raw device chunks, the offset is set as zero.

Regards,
Shriyan

 

[nAtHaNs]

Thanks for all your replies.

In the output of my onstat -d command, I have in the column offset two values, one is 4 and the other is 26000.

What happens if I use a wrong offset value with the command onspaces -s dbs20 -p /dev/data55 -o 16 -O?

Regards.

 

[vpshriyan]

Hi Nathans,

In most cases wherein you provide a wrong offset to a chunk the onspaces utility, refuses the command returning error message as below:

"The chunk specified does not exist."

In some cases, however, if the difference is about 1-3 positive units, it may process your command to put the chunk status to Oline or Down marking.

Regards,
Shriyan