ADSM-L

Re: Performance/striping/etc.

1998-09-03 17:28:08
Subject: Re: Performance/striping/etc.
From: Peter Gathercole <peter.gathercole AT VIRGIN DOT NET>
Date: Thu, 3 Sep 1998 22:28:08 +0100
This appears reasonable. I have achieved similar I/O performance to raw
striped SSA volumes, but I have not been able to achieve anything like this
performance from ADSM.

We have ADSM running as the only traffic over 100Mb/s FDDI, and can achieve
about 6.5MB/s to 3575 tape drives (which is good).

However, using a diskpool striped over 4 SSA disks, I have only been able to
achieve 4-4.5MB/s for the same archive operation.

I am interested in seeing what the bottleneck is. I believe that it must be
in ADSMs disk pool code.

I know that there may be some tuning that can be done regarding the buffer
sizes that ADSM uses to buffer data to the storage media. I have not had the
time to play around with this.

Anybody any comments?

Peter Gathercole
Open Systems Consultant.

Bruce Elrick wrote:

> Given AIX 4.2.1 or higher on an RS/6000 with SSA disk, would the
> following be a reasonable disk allocation policy for performance and
> reliability?  I put my comments after each.  I welcome other comments as
> well.
>
> 1) Use ADSM mirroring for database and recovery log with mirrorwrite set
> to sequential for both
> - Yes, protection from corruption is the highest goal.
>
> 2) Raw logical volumes for all disk volumes (db, log, & stgpool)
> - JFS adds journaling but only for FS structure changes (inode
> allocation, directory structure changes) and thus prvides marginal
> protection.  The primary advantage of JFS is over other non-journaled
> FS's, protecting you from the need to do long fsck's after crashes.
> - Raw LV's waste some space ((PP size - 1MB) for db and log, but this is
> acceptible)
> - JFS adds a (small?) performance hit on top of the LVM overhead.
> - Does JFS add any performance gains via read-ahead and caching that
> aren't realized by raw LV AND are actually useful given the way ADSM
> accesses the volumes (the buffpool and logpool show that the db and log
> is doing its own optimization for disk access)
>
> 3) Use AIX low level striping (software RAID 1) where volume size
> permits
> - I know I can 'dd' (sequential read & write) to SSA striped at close to
> 35 MB/s vs what 7-9 MB/s for single disk access; does this benifit ADSM
> given its access patterns?
> - alternately, does one give one volume to each spindle and let ADSM's
> spreading of load across multiple volumes do the trick?
>
> 4) Put which disks nearest to the SSA adapter card; logs first, then db,
> then stgpool?  Or db, log, then stgpool.
>
> Food for thought...and any input/debate is appreciated.
>
> Bruce
<Prev in Thread] Current Thread [Next in Thread>