Separating reads and writes for performance benefit

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 22 Mar 2012, at 20:40, Haris Zukanovic wrote:

> You do make a nice description of this combination rsync + gluster, but would not it be equal (more or less) to separate reads and writes like I wanted to do? Reads directly from the Gluster brick which is on the local filesystem of the server and writes to Gluster mounted FS for replication to other servers.

Yes, but the performance problems people have reported apparently exist even when using the local-read translator to accelerate them. It's been a while since I benchmarked that, so it might not be the case any more.

I usually run backups directly from the local backing store, bypassing gluster entirely - if I ever need to do a restore, I go in through the proper way to avoid making a mess of any metadata. I guess you could split your reads and writes the same way.

Marcus
-- 
Marcus Bointon
Synchromedia Limited: Creators of http://www.smartmessages.net/
UK info at hand CRM solutions
marcus at synchromedia.co.uk | http://www.synchromedia.co.uk/





[Index of Archives]     [Gluster Development]     [Linux Filesytems Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux