Hi Brian,
We use a 128KB segment size on our DDNs and a 1MB block size and it
works quite well for us (throughput in the 10's of gigabytes per second).
IIRC the sub block (blockSize/32) is the smallest unit of allocatable
disk space. If that's not tuned well to your workload you can end up
with a lot of wasted space on the filesystem.
In option #1, the smallest unit of allocatable space is 256KB. If you
have millions of files that are say 8K in size you can do the math on
lost space.
In option #2, if you're using the same 1MB segment size from the option
1 scenario it gets even worse.
Hope that helps. This might also help
(https://www.ibm.com/support/knowledgecenter/STXKQY_4.2.0/com.ibm.spectrum.scale.v4r2.ins.doc/bl1ins_frags.htm).
-Aaron
On 7/16/16 9:04 PM, Brian Marshall wrote:
All,
When picking blockSize and segmentSize on RAID6 8+2 LUNs, I have see 2
optimal theories.
1) Make blockSize = # Data Disks * segmentSize
e.g. in the RAID6 8+2 case, 8 MB blockSize = 8 * 1 MB segmentSize
This makes sense to me as every GPFS block write is a full stripe write
2) Make blockSize = 32 (number sub blocks) * segmentSize; also make sure
the blockSize is a multiple of #data disks * segmentSize
I don't know enough about GPFS to know how subblocks interact and what
tradeoffs this makes.
Can someone explain (or point to a doc) about sub block mechanics and
when to optimize for that?
Thank you,
Brian Marshall
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at spectrumscale.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss
--
Aaron Knister
NASA Center for Climate Simulation (Code 606.2)
Goddard Space Flight Center
(301) 286-2776
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at spectrumscale.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss
