Alejandro Vargas' Blog

A couple of times I've removed ASM disks from a diskgroup and found that the rebalance operation will take too long.

This is not a problem when you don't want to disturb online users, the rebalance process only locks for writes 1mb of data at a time, so letting the rebalance to run slowly has no performance impact.

My problem was that I needed to free the disks as fast as possible in order to finish some IO tests within a limited timeframe.

The procedure to change the rebalance power of an ongoing operation is very simple:

alter diskgroup DATADG rebalance power 6;

After that the rebalance operation will restart with the new set rebalance power.

For a complete example check this document:

Changing the rebalance power of an ongoing rebalance operation

On Wednesday, July 16 we will have our 3rd RAC and ASM on Linux Forum at the Sheraton Hotel in Tel Aviv, this time dedicated to Performance and Maximum Availability Architecture (MAA).

We have several invited speakers from Oracle Israel: Oded Ilan, Gadi Chen and myself will speak about ASM performance; Eyal Druskin and Nickita Chernovsky will explain the concept of MAA, I'll present a short demo of Active Data Guard.

Then we have two special invited speakers:

Saar Maoz, from Oracle Corporation, Racpack Team will speak about performance tuning of RAC using AWR reports.

Husnu Sensoy, from Turkcell, the main cellphone services provider in Turkey will speak about his experience implementing a 50TB database based on ASM.

The presentations can be downloaded here: Rac on Linux Forum Presentations, If you are based on Israel and didn't register yet please go to this link to see the program and register: Oracle RAC on Linux Forum

See you there!