Forum Discussion

GertjanA's avatar
GertjanA
Moderator
16 years ago

Question on 'splitting' up Journal Store

Hello all,

 

Think with me..

We currently have journalpartitions that are 1 TB in size. Expiry is set for 6 months. general usage per partition is average 85/90%. This obviously is a problem in backing these up. Talking with the backup team, they advised they can back this up more easily (and quicker) if I was to use a size of 250/300 GB per partition. I'm thinking of reconfiguring, so that a journalpartition is 250 GB. Diskwise, storage will provide/present 3 or 4 disks to the journalarchivingservers so I can easily close a partition, and open a new partition on a new disk. Obviously, the expiry scheme stays the same. The indexes are no problem.

 

I need to figure out how long it takes to fill 250 GB. Anyone knows of a SQL query I can run? Any other idea to keep this in manageable size? I know EV8 has automatic rollover based on size/date, but we're not upgrading soon.

 

Thanks for thinking with me.

Gertjan

 

  • This might help you out.  just put in your date range and see daily rate with total size 

    -- This one gives a daily rate:

    --Runs against the VaultStore Database

    select "Archived Date" = left (convert (varchar, archiveddate,20),10),

    "Daily Rate" = count (*),

    "Av Size" = sum (itemsize)/count (*),

    "Total Size" = sum (itemsize)

    from saveset

    where archiveddate between '2004-05-01' and '2009-05-31'

    group by left (convert (varchar, archiveddate,20),10)

    order by "Archived Date" desc

4 Replies

  • This might help you out.  just put in your date range and see daily rate with total size 

    -- This one gives a daily rate:

    --Runs against the VaultStore Database

    select "Archived Date" = left (convert (varchar, archiveddate,20),10),

    "Daily Rate" = count (*),

    "Av Size" = sum (itemsize)/count (*),

    "Total Size" = sum (itemsize)

    from saveset

    where archiveddate between '2004-05-01' and '2009-05-31'

    group by left (convert (varchar, archiveddate,20),10)

    order by "Archived Date" desc

  • You could also look at your backups for the last month and see what is the relative growth each day and just do some extrapolation of the figures. 

     

    I think the script gives the size of the item before archiving therefore it is not a true reflection on what size the actual vault store partition is.  Therefore, you might think that you are up to 250GB more quickly.

    This being said I still like the script idea as we are not talking precise facts and figures!

  • That is right, the item size is the original size before compression.  Sorry I forgot to mention that earlier.  :)
  • ahh. I was not clear again.. It is not being backed up, so cannot check that. sorry. But, the script will do! I'm trying to get an estimate, and using the script, I can at least sort of calculate dailey/weekly/monthly additions.

     

    Thanks guys.