Date copy rate is slower in from a flexclone

[deygaurab]

Hi Friends,

I have created a flexclone volume of my production database. The DBA's are copying some files from the cloned volume/luns. The are complaining about slow performace 3 Mb/s data rate while copying some files from cloned volumes to destination(the luns of which happens to be on the same controller)

Does anyone have any similar experiences with the cloned volume?

Thank you

Rahul

[radek_kubka]

Are you saying copying from a clone is slower than copying from the original volume?

In other words: are we sure cloning is to be blamed, rather than it is slow for some other reasons?

[deygaurab]

Hi Radek,

The clone is presented to a server(AIX Lpar) where there is already an exisiting fucntional database. The motive is to copy from files from the restored LUNs to the existing datbase.To facilitiate this, I cloned the volume and mapped the LUNs to the AIX Lparserver.  Once the LUNs were mounted, the DBA's started the copying files (inside some mount points) from the cloned volumes(mapped LUNs)  to a datbase which is on the same AIX lpar server.  The hosts and the storage are on the same data center and connected to the same switch. I am finding it difficult to understand why the rate of data copy is so slow.

Regards

Rahul

[deygaurab · Accepted solution]

I found the solution. Changing the mount point option worked for me. I needed to change CIO option in the mount point and now it works 10 times faster.

Cheers

Rahul

[scottgelb]

Concurrent I/O is really nice and I remember reading about it on aix several years ago.

The production mount was CIO and the clone wasn't ? Multiple interfaces used too?