cancel
Showing results for 
Search instead for 
Did you mean: 

Disk storage unit full

philalbert
Level 4

My disk storage unit is getting full when full backups are running on the weekend and I'm not sure how to make it so that it doesn't happen.

Here are my current settings:

Maximum concurrent jobs: 20

High Water mark: 60 %

Low water mark: 15 %

Enable Temporary Staging Area is activated.

 

Why is my volume getting full and not emptying fast enough for the new data?

 

Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
More disk space is what you needed. Your HWM is also too low. You will get disk full error when disk space gets to 60% . The default HWM is something like 98% . Why was this lowered? If disk size is right and HWM not too low, you should have a good couple of days' backups on disk that will give you fast restores (as disk backups was intended for ).

View solution in original post

12 REPLIES 12

nbutech
Level 6
Accredited Certified

You will need to manually run nbdelete -allvolumes which run every 12 hrs

 

Once the duplication is complete the disk space is not cleared right away

There are 2 steps after that image expiration and actual disk space deletion

Once that is done NBU master will pass the message to OS to clear the disk space

You can also check admin logs if nbdelete has run 

philalbert
Level 4

In the admin log I can see that the command nbdelete has ran multiple times before I got the disk full error.

Maybe there's just to much data trying to get copied in the job, I'll take a look. (Last time I check there was enough space.)

 

emret
Level 6
Partner Accredited

is your storage unit msdp or adv disk. what type of backup you are taking ? if you are using compression enable backup for database level backup it may cause that problem

sdo
Moderator
Moderator
Partner    VIP    Certified

Size of volume?

Total size of one set of weekly backups?

What is the second stage of the DSSU?

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
This sounds to me like BasicDisk, right? And you mentioned 'temporary staging'. Hopefully this means staging to tape? Can you check the staging schedule to see how often the staging is taking place? Can you see duplication jobs running successfully? NBU Admin Guide I recommends that BasicDisk size should be big enough to hold at least 2 full backups. Can you confirm that yours is big enough? When all of the above have been checked and confirmed, there are TWO things that determine image lifespan on disk: 1) Successful staging to tape. Only images successfully duplicated to tape will be expired/deleted when HWM is reached. 2) Retention level in backup schedule. If disk backups have not been staged to tape, then the retention level/period in policy schedule is applied.

Nicolai
Moderator
Moderator
Partner    VIP   

Stage to tape should not happen at the same time as backups are written.

Backup at night, stage to tape during day time.

Ensure NUMBER_DATA_BUFFERS & SIZE_DATA_BUFFERS are configured in Netbackup, this is important for tape performance.

http://www.veritas.com/docs/000094646

Usage tease values :

SIZE_DATA_BUFFERS = 262144

NUMBER_DATA_BUFFERS = 256

 

philalbert
Level 4

The Disk type is BasicDisk. 
We are not using compression.

Yes staging is to tape.

Disk Staging schedule:
Priority: 99
Frequency: 1 days
Start Window: Everyday 6 AM to 8 PM.

Yes Duplication jobs are running successfully every day.

The buffers are configured correctly. 


There's not enough space to hold 2 Full backups of all the servers from the policies writing to that Storage unit.
They do not run all at the same time though, maybe that's why I don't always get the not enough space error.

philalbert
Level 4

I've added more space to the disk. The backup was probably writing to much data or the disk was getting full to fast.

I'll let the backup run for a week or two and see how it's going.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
More disk space is what you needed. Your HWM is also too low. You will get disk full error when disk space gets to 60% . The default HWM is something like 98% . Why was this lowered? If disk size is right and HWM not too low, you should have a good couple of days' backups on disk that will give you fast restores (as disk backups was intended for ).

philalbert
Level 4

Looks like my understanding of the HWM was not accurate. I didn't know that I would get a disk full error if the HWM is reached. I thought the backup to tape would get started sooner if the HWM was lower.   :\

I'll set it back at 90% where it was before. (Or should I put it at 98% like you are suggesting?)

 

Thanks Marianne.

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Duplication to tape will only take place during the duplication schedule that you have configured. HWM of 98% is good IMHO.

philalbert
Level 4

Ok, I'll change it.

 

Thanks