After Lvmerge more backup time

[SR_2006]

Dear all,
After splitting Mirror LV we r taking backup of one of the split LV.But now backup time reqd. for LV is incresed & backup progress is very slow suddenely.other backup is working OK.We are using 2 node cluster.
thnks in advance.

srl

[Steven E. Protter]

Shalom,

I'm suggesting a $1,100 per CPU(your price may vary) add in product that can do this quickly.

OnlineJFS.

OnlineJFS enables the following process.

1) Down the database for 1 minute
2) Create a tempory write queue logical volume
3) Restart the database with all writes going to this special lv you just created on the fly.

The database continues to run and you can take a cold OS backup of the sql database and it really does not matter how long it takes.

4) Online JFS command merges the writes with the original logical volume and then the temporary LV can be removed.

It involves less than a minutes downtime and works very, very nicely.

SEP

[SR_2006]

thanks steven,
It was noticed that one HDD of array was gone failure & array was in rebuild state in whole night.when disk gone bad how much time array takes to rebuild?

thanks.

[Patrick Wallek · Accepted solution]

If the array is in a rebuild state, then it is entirely possible that your I/O performance will be impacted.

The time it takes for the array to rebuild depends on how large the drive is, how fast the array controllers are and the rebuild priority of the array.

If rebuild priority is set very low, which would have the least impact on array performance, then the rebuild could take many many many hours. If the rebuild priority is set high, which would have the most performance impact, the rebuild time could be significantly less.

[SR_2006]

Thanks patrick,
How to set the rebuild priority low of an array?Also now I have replaced the faulty disk with new one now array is in optimising state but now system performance is not degraded.what is technical reason for this?
thanks.