Re: Unlikely performance issues

suffering a write penalty because of the use of RAID 5 which will
certainly slow things done.
Well, raid 5 is very fast with read operations. I'm quit sure I am having
troubles with reading data. That's what, in my opinion, is the odd thing.

An ideal disk layout would be mirrored system
drives and separate RAID 10 drives for databases and logs.

I don't agree. Ideal would be RAID 5 for the databases, raid 10 for the
logs, raid 1 for the pagefile. But I still will insist that 50 users with
quite large databases shouldn't hurt performance so much as it does. I'm
quite sure something else is wrong.




"Nuevo" wrote:

You are suffering a write penalty because of the use of RAID 5 which will
certainly slow things done. An ideal disk layout would be mirrored system
drives and separate RAID 10 drives for databases and logs. However, if you
do not have the ability to do this I would think about moving the pagefile,
and search catalog to C and put the rest on D. See if this helps. I'd also
run diagnostics on the disk and controller just to be sure.

Nue


"Nico" <Nico@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:C5F5076F-4C00-40FB-A53F-512A111D8626@xxxxxxxxxxxxxxxx
Hi,

We run exchange 2003 (standard) / windows 2003 (standard). The server has
the following disk configuration:
C: scsi raid 5
D: ide raid 1

c: has the transaction logs + mailbox databases
d: has the search catalog + paging file + public databases

the server has 4GB ram and a xeon 2.4GHz cpu

the exchange server has 50 mailboxes. Avarage mailbox has 7000 items and
is
about 300 MB. some mailboxes have over 50.000 items and some are larger
than
3GB.

Users have performance issues (outlook reports non responsive exchange
server).

I understand that best performance is reached by having the databases on
seperate disks, and having the logfiles on seperate disks. In my opinion
this
shouldn't be necessary on a server with just 50 clients. I don't
understand
that an exchange server, wich should be able to handle 1000ths of clients
is
having troubles with just 50 of them.

I noticed the following in the performance analyzer:
disk queue/disk time of C: reaches 100% on a simple sort operation in
outlook (size) on my own inbox (10.000 items)
pages/sec has an avarage of 0.2
% processor time has an avarage of 3%
The exchange server performance troubleshooting analyzer only indicates is
should move my transaction logs to another disk.

So, thing point out that I should move my transaction logs. I just can't
accept I need more/better hardware for just those few users. Okay okay,
our
.edb file = 30GB and the .STM = 4 GB.. Is this causing all the trouble?

Any comments are very welcome. Thank you,

Nico Venema



.

Relevant Pages

  • slackware 9.1 software raid problem
    ... Setting up a RAID system with Slackware 8 is not extremely difficult once ... mirroring the root partition and booting from that mirror was not possible. ... Each disk is attached to a different IDE chain on the motherboard. ... The ability to boot from the Slackware 8 install CD. ...
    (alt.os.linux)
  • Re: Disk Performance issues?
    ... All our DB are on a single partition that lives on the single virtual disk ... made up of the RAID 10. ... LUN ... your databases on the single RAID 10 ...
    (microsoft.public.exchange.admin)
  • Re: Is it HighPoint, is it Seagate, or is it Windows 2000 Professional
    ... I have really learnt about the uses and limits of RAID. ... Use all the manufacturer recommended system recovery ... provisions -- Emergency Disk, have a DOS boot disk ready, ... what could cause Windows to suddenly change its mind ...
    (microsoft.public.win2000.setup)
  • Re: [opensuse] 10.2 no RAID to 11.0 RAID 1
    ... 10.2 no RAID to 11.0 RAID 1 ... you can force rebuilds in mdadm in situations where no firmware raid will ever let you. ... Send me your problem disks that you think are impossible to assemble and I bet in a little while I can tell you how to assemble the array as long as there actually is enough there to use. ... One thing I've seen, which I don't think is your problem but shows the kind of thing that happens, a disk will drop out of the array and reappear instantly as a higher drive letter than the system really has. ...
    (SuSE)
  • LVM on SW RAID for sarge - success
    ... I've just had success migrating Debian Sarge to root on LVM on RAID. ... First make sure the SCSI BIOS boots from the first disk. ... I made reiserfs on all the filesystem partitions and completed the ... The remaining two disks were partitioned and used to create RAID arrays ...
    (Debian-User)
Loading