Filesystem performance

Hello

I am installing a new zimbra server and I am following this guide to improve
the performance of the server

Performance Tuning Guidelines for Large Deployments - Zimbra :: Wiki

In the File System section it suggests to use the following parameters

-j Create the file system with an ext3 journal.
-L SOME_LABEL Create a new volume label. Refer to the labels in /etc/fstab
-O dir_index Use hashed b-trees to speed up lookups in large directories.
-m 2 Only 2% needs to be reserved for root on large filesystems.
-i 10240 For message store, option -i should be the expected average message size. Estimate this conservatively, as no. of inodes can not be changed after creation.
-J size=400 Create a large journal.
-b 4096 Block size in bytes.
-R stride=16 Stride is used to tell the file system about the size of the RAID configuration. Stride * block size should be equal to RAID stripe size. For example 4k blocks, 128k RAID stripes would set stride=32.

What I found is that with the above configuration for the filesystem
if I run the "bonnie++" benchmark tool in it, it is much slower than with the
default parameters created by the Fedora 7 installer

Here is the result of bonnie++ in this partition with the above parameters

Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
s2 4G 36591 9 17650 3 42325 4 106.7 0
------Sequential Create------ --------Random Create--------
-Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
s2,4G,,,36591,9,17650,3,,,42325,4,106.7,0,16,+++++ ,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++,+++++ ,+++

36591 write speed = 36MB/s


Here is the result of bonnie++ in a partition formatted by fedora

Version 1.03 ------Sequential Output------ --Sequential Input- --Random-
-Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP
s2 4G 59103 15 26123 5 67176 7 122.6 0
------Sequential Create------ --------Random Create--------
-Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
s2,4G,,,59103,15,26123,5,,,67176,7,122.6,0,16,++++ +,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++,++++ +,+++

59103 write speed = 59MB/s

The difference is 36MB/s with ext3 formatted with zimbra parameters and
59MB/s formatted by default with fedora (write performance)


So the question is if those parameters are really important for zimbra
or if I should leave the partition formatted by fedora as it gives better
performance in the bonnie++ benchmark?

Thanks

Other factors need to be considered, not just the pure throughput. Number of files, average size of files, interleaving, type of underlying storage, RAID, block sizes etc etc.

Ok, but in our case the hardware and the raid wont change, we just need
to decide the best way to format the partition, with the Zimbra parameters
or with the Fedora parameters.

Avarage number of messages will be 10.000 per mailbox and avarage message size 15KB.

I will try to find which of the zimbra parameters makes the bonnie benchmark much slower.

[]s