Jump to content

Unbelievable search times with similar file names


Recommended Posts

I am using Retrospect 7.0 with DLT-v4 drives on a dedicated Backup-Server (Dell 2*1000 MHz. 2 GB RAM) to save a Cyrus IMAP Mailstore. The mails in the Mailstore are files with a number as name, mailfolders have user-defined names. There are 190,411 files and 1,718 folders. This is a directory listing:

 

…411 files before...

a/user/annett/53131.

a/user/annett/53132.

a/user/annett/515.

a/user/annett/516.

a/user/annett/51695.

a/user/annett/517.

a/user/annett/518.

a/user/annett/519.

a/user/annett/51911.

…189991 more…

 

Each attempt to recover a file of a storage set which contains the mailstore takes more than 4 hours (at full CPU load). This time needs Retrospect to scan the storage set, not to execute the recovery process!

 

To the comparison: the storage set contains 4 Mio. files of other servers, too. A scan of this files takes less than one minute!

 

I think, this is a balancing-problem of the B-Tree of the database index.

But what can I do?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...