Archive Roll over

[reachmitra]

Hi,

Is this a good idea to create new archive and roll over to new archive quarterly or once in every six months instead of having a large archive?Does symantec recommends this? What are the benifits interms of performnace improvement in EV,CA and DA?

Will there any improvement in discovery accelerator search/export performance?

Think of two scenario . 

1.I have a large archive A which contains data from 2008 to 2014.I am using date range  2010 to 2013 as search criteria

2. I am creating archive yearly. so 2008_A,2009_A, 2010_A , 20011_A, 2012_A, 2013_A & 2014_A . I am using same date range 2010 to 2013 and selecting 2010_A,2011_A,2012_A & 2013_A archives in DA case prperties

Which search will be faster?

Similarly is there any benifit in creatng vault store once in every six month to keep the size small?

One improvement I can think of is index repair/rebuild will take less time.

Waiting for your valuable inputs?

[JesusWept3 · Accepted solution]

Most really big companies i've worked with do exactly like you said
so they will have like

JournalArchive_2013_Qtr1
JournalArchive_2013_Qtr2
JournalArchive_2013_Qtr3
JournalArchive_2013_Qtr4
JournalArchive_2014_Qtr1

There are multiple benefits, many you mentioned
- Index Rebuilds/Repairs/Conversions are easier/quicker when its limited to a short time range
- EVSVR Verify/Repairs can be quicker when limiting it to a certain archive
- DA Searches can be more precise, and i've seen several smaller archives search quicker than one monster archive
- If you reach the 2 Billion limit in a vault store, you can easier manage the archives when new vault stores need to be created
 

[Nate_D1]

Another bonus is quicker backups which is really nice for environments with small backup windows.

We also do quarterly archives for both journal and mail seperately. We decided to do our archives as mount points underneath C: after realizing we would run out of drive letters pretty quickly! Just a few more thoughts.

[EV_Ajay]

Hi,

In DA if you create small archives the DA search became faster. 

In DA it will never look for the archive which doesn't come in search criteria date range. 

The DA search and export performance will increase.