Answer the question
In order to leave comments, you need to log in
SSD testing and results
Good afternoon. Ran into strange results when testing dedicated servers.
Servers:
1) MSI H87-G43 motherboard (6 channels SATA 6GB/s). 4 drives in softraid 10: 2 Samsung SM843T drives and 2 Intel SSD 520 series 240 Gb drives.
2) MSI H87-G43 motherboard (6 SATA 6GB/s channels), but it costs 2 Intel SSD 520 series 240 Gb disks in soft raid1.
3) Mother - SuperMicro X9SCMF + dedicated controller LSI_MegaRAID_SAS_92404i. 4 hard raid 10 disks: 2 pcs. PNY_Prevail_Elite and 2pcs. PNY_Prevail_5K for 240 Gb.
4) Mother - SuperMicro X9SRI3F. 4 soft raid 10 disks: 4 pcs. Kingston_SSDNow_KC300 480 GB.
dd utility results on centos:
Writing (dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync oflag=direct)
1) 366
2) 350
3) 225
4) 237
Read (dd if=test of=/dev/null bs=1M count=1024 iflag=direct)
1) 886
2) 392
3) 721
4) 523 The
question is why 3rd server reading 1.5-2 times less than on the first two configs? Who is to blame (motherboard/controller/drives)? Maybe I'm testing/compiling something incorrectly?
Answer the question
In order to leave comments, you need to log in
IMHO. Perhaps because PNY itself is slower than Intel will be ....
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question