Sorry for the crossposting, i read pots on this forum and I wasn't able to find a solution.
Actually we are running a linux server with multipath. MD3200i are directly connected whit 8 1Gb Ethernet card.
here the multipah conf
defaults {
udev_dir /dev
polling_interval 10
selector "round-robin 0"
path_grouping_policy multibus
getuid_callout "/lib/udev/scsi_id --whitelisted --device=/dev/%n"
prio const
path_checker directio
rr_min_io 3
rr_min_io_rq 5
flush_on_last_del no
max_fds 8192
rr_weight priorities
failback immediate
no_path_retry fail
queue_without_daemon no
user_friendly_names no
mode 644
uid 0
gid disk
}
blacklist {
wwid 26353900f02796769
devnode "^(ram|raw|loop|fd|md|dm-|sr|scd|st)[0-9]*"
devnode "^hd[a-z][[0-9]*]"
wwid 36848f690ea06050018916f2817c3b9f9 #sda
wwid 36848f690ea06050018916f2a17eac886 #sdb
device {
vendor DEC.*
product MSA[15]00
}
}
blacklist_exceptions {
devnode "^dasd[c-d]+[0-9]*"
wwid "IBM.75000000092461.4d00.34"
}
devices {
device {
vendor "DELL"
product "MD32xxi"
path_grouping_policy group_by_prio
prio rdac
polling_interval 5
path_checker rdac
path_selector "round-robin 0"
hardware_handler "1 rdac"
failback immediate
features "2 pg_init_retries 50"
no_path_retry 30
rr_min_io 3
rr_min_io_rq 5
prio_callout "/sbin/mpath_prio_rdac /dev/%n"
}
}
multipath -ll
3690b11c0002c79dc000003515119ded4 dm-0 DELL,MD32xxi
size=36T features='3 queue_if_no_path pg_init_retries 50' hwhandler='1 rdac' wp=rw
|-+- policy='round-robin 0' prio=6 status=active
| |- 7:0:0:0 sdc 8:32 active ready running
| |- 9:0:0:0 sdd 8:48 active ready running
| |- 8:0:0:0 sdf 8:80 active ready running
| `- 14:0:0:0 sdj 8:144 active ready running
`-+- policy='round-robin 0' prio=1 status=enabled
|- 11:0:0:0 sde 8:64 active ghost running
|- 12:0:0:0 sdg 8:96 active ghost running
|- 10:0:0:0 sdh 8:112 active ghost running
`- 13:0:0:0 sdi 8:128 active ghost running
A bref speed test show
d of=dd2.dat if=/dev/zero bs=1G count=65
65+0 records in
65+0 records out
69793218560 bytes (70 GB) copied, 131.394 s, 531 MB/s
dd if=dd2.dat of=/dev/null
136314880+0 records in
136314880+0 records out
69793218560 bytes (70 GB) copied, 434.808 s, 161 MB/s
I played with MTU disabling/enabling JUMBO frames and I tried various size of MTU.
I also played with reading cache sizes and flush settings as stated in other posts. Improving a little bit performances but the write/read ratio constantly remain about 4:1.
Any suggestion will be appreciated.
Bests