What is Oracle ASM scrubbing? Oracle ASM scrubbing is a feature that enhances availability and reliability by seeking out rarely accessed the data and fixing logical corruptions automatically.
You can perform on-demand based scrubbing by using altered disk group scrub command. And you can run this command at ASM disk group level by running all the disk group and scrub option. And it is also possible to scrub at the disk level. The second command you see the scrub disk and the disk name to specify the disk– will be scanned to identify logical corruptions.
And it’s also possible to scrub at the file level, like a third example. All the disk group data scrub file, and then this is to scrub the specified file. And we can also add additional options along with the scrub option.
If you want to stop ongoing scrub operation, you can run all the disk group with a scrub stop option. And this feature was available prior to 23ai, so what you knew in 23ai. With the introduction of Oracle Database 23ai is now possible to scrub individual extents or a range of extent.
Previously, ASM scrubbing was only available on file levels or disk level and disk group level. When compared to the scrubbing the entire file in 23ai, you can specify extent. You can scrub specific extent set to reduce scrubbing turnaround time. It improves the data availability and also minimizes the performance impact.
It’s possible to perform the extent-based scrubbing by using same command alter disk group scrub, but with the additional option. So when you look at the first command all to disk group scrub file and block number 10 and count to three. So what it does is this command identifies block number 10. And from that, check the three blocks. So block number 10 and 11 and 12.
And identify extent that contain data blocks and then scrub. So the extent-based scrubbing instead of scrubbing entire file, we can specify specific blocks. Then the command automatically identify extent that contain blocks you specified. So it will dramatically reduce the turnaround time. It will also possible to scrub multiple groups of a data blocks.
So you can add a block and count the combination multiple times. So in the example, in the second bullet, we specify the block number 50 and count to 10 and block number 1,024 count to 70, meaning that identify block number 50 and then count 10 blocks from that. And then identify block number 1,024 and identify 70 blocks from the block number 1,024.
And then once blocks are identified and check the extent that contain data blocks. So extent-based scrubbing will be performed based on those extent that contain blocks you specified. So it dramatically improves the performance.
Recent Comments