Re: 4x lower IOPS: Linux MD vs indiv. devices - why?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Andrey,

Am 23.01.2017 um 20:06 schrieb Kudryavtsev, Andrey O:
Hi Tobias,
Yes, “imsm” is in generic release, you don’t need to go to the latest or special build then if you want to stay compliant. It’s mainly a different layout of a raid metadata.

Your findings follow my expectations, for QD1 sync engine does good results. Can you try libio with QD4 and 2800/4 jobs?
Most of the time I’m running Centos7 either with 3.10 or latest kernel depends of the scope of the testing.

Changing sector to 4k is easy, this can really help. see DCT manual, it’s there.
This can be relevant for you https://itpeernetwork.intel.com/how-to-configure-oracle-redo-on-the-intel-pcie-ssd-dc-p3700/



I have gone through the whole manual, but I cannot find info about the meaning of different LBAFormats.

The Oracle article above uses

LBAFormat=3

which I presume means 4k secor size.

The P3608 seams to support a value up to 6:

oberstet@svr-psql19:~/scm/parcit/RA/user/oberstet$ sudo isdct show -all -intelssd 0 | grep LBA
LBAFormat : 0
MaximumLBA : 3907029167
NativeMaxLBA : 3907029167
NumLBAFormats : 6

So is this the correct mapping for the value?

LBAFormat	Sector Size
0	512
1	1024
2	2048
3	4096
4	8192
5	16384
6	32768

In this case, I'd use

LBAFormat=4

to get 8k sectors, sine my workload is purely 8k.

Cheers,
/Tobias

--
To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [Linux Kernel]     [Linux SCSI]     [Linux IDE]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux SCSI]

  Powered by Linux