Forum Discussion

halq's avatar
halq
Level 3
5 months ago
Solved

how long the pure disk pool to reflect the new size after did "expire" the backup image manually.

Hi folks,

We have a Veritas NetBackup version 10.0.0.1, due to the pure disk pool size is increasing a lot and this is not the time for expansion, so I did "expire" some catalog backups to free up the disk pool size.
I did expire a lot of old images (~20%), but the disk pool is not reflect the new size. Can you please let me know how long it takes for the disk pool to reflect the new size?

Thank you so much
Ha

  • Hey!

    Well this shouldn't take longer than 24h, until you will expedite it. To expedite it do the following:

    on primary: bpimage -cleanup -allclients

    on storage server holding MSDP

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueue

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueue

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueueinfo

    sources:

    https://www.veritas.com/support/en_US/article.100000714

    processqueue has to be ran twice, and once processqueueinfo will show no for busy and pending some space should be returned back, but I would not expected a lot as all backups are deduped and 20% will not produce hundreds of GB/TB 

3 Replies

  • Hey!

    Well this shouldn't take longer than 24h, until you will expedite it. To expedite it do the following:

    on primary: bpimage -cleanup -allclients

    on storage server holding MSDP

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueue

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueue

    /usr/openv/pdde/pdcr/bin/crcontrol --processqueueinfo

    sources:

    https://www.veritas.com/support/en_US/article.100000714

    processqueue has to be ran twice, and once processqueueinfo will show no for busy and pending some space should be returned back, but I would not expected a lot as all backups are deduped and 20% will not produce hundreds of GB/TB 

    • halq's avatar
      halq
      Level 3

      Thanks Quebek, I followed your suggestion after revise the article.714. It's worked well.

      And yes, the free up space is not equal to the data record removed by deduped rate, just for my case is ~63% of the total size force "expired"

  • Quebek provides the procedure to manually start the garbage collection of the MSDP storage.
    However, the results may not be what you expect. Deduplication works by saving only the new blocks and linking the old blocks to the new backup.

    So, if the backups you deleted had a 99% deduplication ratio, you should not expect to free up significant space.