Hello,
I'm doing some research on serverless backup and was hoping somebody could shed some light on the topic for me.
All the research papers that I've found talk about the server doing a snap shot of the data, then sending an Extended Copy command to the data mover with the starting location of the snap shot and how many blocks to back up.
So with this I'm assuming the Master Backup server keeps the schedule of backups, file indexes and databases. When the Master Server (MS) finds that it needs to backup an Application Server (AS) it sends a notice to the AS serverless backup agent stating that it has a backup scheduled. At this point the AS makes a snap shot of the data to be backed up and then sends the pointer information, i.e. location and blocks, back to the MS which then forwards that information to the data mover. The data mover then uses the extended copy command to retrieve the data from the disk on the AS and streams it to tape (if you are using tape). When complete, the data mover lets the MS know the job is complete with no failures and continues on to the next AS.
Is this a correct scenario?
Now, another question I have is if the AS has to release it's drive so the data mover can mount it to get access to the information? In this scenerio then the AS will experience downtime while the backup is taking place.
or
Can the data mover do a low level block copy using Extended Copy without mounting the device?
Thanks for any help..
I'm doing some research on serverless backup and was hoping somebody could shed some light on the topic for me.
All the research papers that I've found talk about the server doing a snap shot of the data, then sending an Extended Copy command to the data mover with the starting location of the snap shot and how many blocks to back up.
So with this I'm assuming the Master Backup server keeps the schedule of backups, file indexes and databases. When the Master Server (MS) finds that it needs to backup an Application Server (AS) it sends a notice to the AS serverless backup agent stating that it has a backup scheduled. At this point the AS makes a snap shot of the data to be backed up and then sends the pointer information, i.e. location and blocks, back to the MS which then forwards that information to the data mover. The data mover then uses the extended copy command to retrieve the data from the disk on the AS and streams it to tape (if you are using tape). When complete, the data mover lets the MS know the job is complete with no failures and continues on to the next AS.
Is this a correct scenario?
Now, another question I have is if the AS has to release it's drive so the data mover can mount it to get access to the information? In this scenerio then the AS will experience downtime while the backup is taking place.
or
Can the data mover do a low level block copy using Extended Copy without mounting the device?
Thanks for any help..
