A couple of things: the cache hit % is too low. we'd like to see >99%
ideally, but 90% is bad performance. and cache wait % should be 0
ideally.one reason could be the many volumes you are using, and the fact
that Veritas filesystem does introduce an overhead.
bufpoolsize should be normally around 10% of your real RAM, but if
increasing it increases paging on the server then don't.
the DB, being read oriented, performs better if it is split among different
physical volumes. but not more than 10-12 volumes.together with what Zlatko
said about arrays.
and yes, switch from RAID 5 to RAID 10 and use DB mirroring.that is,
mirrorwrite db parallel and database page shadow.
have fun.
p.s. how's JPL doing? I have a monstrous respect for you guys.
Cordiali saluti
Gianluca Mariani
Tivoli TSM Global Response Team, Roma
Via Sciangai 53, Roma
phones : +39(0)659664598
+393351270554 (mobile)
gianluca_mariani AT it.ibm DOT com
----------------------------------------------------------------------------------------------------
"The people of Krikkit,are, well, you know, they're just a bunch of real
sweet guys, you know, who just happen to want to kill everybody. Hell, I
feel the same way some mornings..."
"Kelly J. Lipp"
<[email protected]
om> To
Sent by: "ADSM: ADSM-L AT VM.MARIST DOT EDU
Dist Stor cc
Manager"
<ADSM-L AT VM DOT MARI
bcc
ST.EDU>
Subject
Re: Database Questions
21/11/2002
00.01
Please respond
to lipp
You might also set the selftunebuffpoolsize yes as well. Check the exact
syntax, but this relatively new option will have TSM adjust this parameter
is necessary.
Also, I like to have this parameter set to some power of two rather than
just a number. Don't know if that matters, but it appeals to my digital
sense.
Kelly J. Lipp
STORServer, Inc.
485-B Elkton Drive
Colorado Springs, CO 80907
lipp AT storsol DOT com or kelly.lipp AT storserver DOT com
www.storsol.com or www.storserver.com
(719)531-5926
Fax: (240)539-7175
-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU]On Behalf Of
Luke Dahl
Sent: Wednesday, November 20, 2002 3:25 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: Database Questions
Output from q db f=d:
Available Space (MB): 50,012
Assigned Capacity (MB): 45,012
Maximum Extension (MB): 5,000
Maximum Reduction (MB): 10,720
Page Size (bytes): 4,096
Total Usable Pages: 11,523,072
Used Pages: 7,574,343
Pct Util: 65.7
Max. Pct Util: 76.2
Physical Volumes: 51
Buffer Pool Pages: 10,500
Total Buffer Requests: 180,673,719
Cache Hit Pct.: 90.08
Cache Wait Pct.: 1.74
Backup in Progress?: Yes
Type of Backup In Progress: Full
Incrementals Since Last Full: 0
Changed Since Last Backup (MB): 64.07
BufPoolSize 42000
I increased the bufpoolsize from 37000 to 42000 a few days ago to bring the
ratio up? Will this affect performance on the system (swap usage or
paging)?
Yes, DB volumes are on Veritas filesystem... I can get you the exact
Veritas
levels if it would help...
Thank you very much, I really appreciate the help!
Zlatko Krastev/ACIT wrote:
> You can find long discussions on this topic in the list archives.
> - it is mostly disadvantageous to have more than one or two DB volumes
per
> disk/array - parallelism you create with more volumes results disk heads
> moving back and forth. You are shooting yourself in the leg.
> - RAID 5 is definitely not very good for TSM DB and for average or
> heavy-loaded server might be disastrous for performance. For small
servers
> might be just fine. Your server with 35 GB DB does not fit in second
> category.
> - "sessions running for hours" sounds terrible. What is DB cache hit
> ratio? Do you have DB volumes on Veritas filesystem?!?
>
> Zlatko Krastev
> IT Consultant
>
> Luke Dahl <ldahl AT JPL.NASA DOT GOV>
> Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
> 07.11.2002 19:47
> Please respond to "ADSM: Dist Stor Manager"
>
>
> To: ADSM-L AT VM.MARIST DOT EDU
> cc:
> Subject: Database Questions
>
> Does anyone know of any advantage/disadvantage of the file sizes for the
> database? Is there an advantage to creating many 1Gb .db files over
> fewer 10Gb .db files? Also, we're running TSM 4.2.1.15 on Solaris 5.8
> using raid 5. I've heard performance can be much greater with raid 0.
> Any truth to that? We're seeing load averages above 10 nearly every day
> and TSM performance is pretty poor. Our database size is 35Gb and
> sessions are running for hours (even small incrementals of various
> workstations). Network bandwidth hasn't peaked over 50% in any 24 hour
> duration. Any thoughts? Many thanks in advance.
>
> Luke Dahl
> NASA - Jet Propulsion Laboratory
> 818-354-7117
|