Is there a guide of how to set multipath for petasan iscsi disk in linux?
shadowlin
67 Posts
April 18, 2018, 6:23 amQuote from shadowlin on April 18, 2018, 6:23 amI tired to follow the guild from lrbd wiki but It doesn't work out.The documentation sector of petasan has only guide for windows.
I tired to follow the guild from lrbd wiki but It doesn't work out.The documentation sector of petasan has only guide for windows.
shadowlin
67 Posts
April 18, 2018, 7:38 amQuote from shadowlin on April 18, 2018, 7:38 amI figured out the problem was the vendor name.
after using this /etc/multipath.conf setting it works
defaults {
user_friendly_names yes
}
devices {
device {
vendor "(LIO-ORG|PETASAN)"
product "RBD"
path_grouping_policy "multibus"
path_checker "tur"
features "0"
hardware_handler "1 alua"
prio "alua"
failback "immediate"
rr_weight "uniform"
no_path_retry 12
rr_min_io 100
}
}
Is there any other setting should be take care of?
I also notice dmesg info of following:
[ 617.801221] scsi 7:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.802071] scsi 8:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.838207] sd 7:0:0:0: Attached scsi generic sg1 type 0
[ 617.840390] sd 7:0:0:0: [sdb] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.840491] sd 7:0:0:0: [sdb] Write Protect is off
[ 617.840497] sd 7:0:0:0: [sdb] Mode Sense: 43 00 00 08
[ 617.840722] sd 7:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.851077] sd 7:0:0:0: [sdb] Attached SCSI disk
[ 617.857993] sd 8:0:0:0: Attached scsi generic sg2 type 0
[ 617.859795] sd 8:0:0:0: [sdc] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.859914] sd 8:0:0:0: [sdc] Write Protect is off
[ 617.859917] sd 8:0:0:0: [sdc] Mode Sense: 43 00 00 08
[ 617.860114] sd 8:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.871043] sd 8:0:0:0: [sdc] Attached SCSI disk
[ 618.546081] device-mapper: multipath round-robin: version 1.2.0 loaded
[ 618.554183] alua: device handler registered
[ 618.554504] sd 8:0:0:0: alua: supports implicit TPGS
[ 618.554511] sd 8:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 2
[ 618.554516] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.570041] sd 8:0:0:0: alua: transition timeout set to 60 seconds
[ 618.570049] sd 8:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.616605] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.616642] sd 7:0:0:0: alua: supports implicit TPGS
[ 618.616650] sd 7:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 1
[ 618.616654] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.630160] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.630286] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
Is there anything wrong?
I figured out the problem was the vendor name.
after using this /etc/multipath.conf setting it works
defaults {
user_friendly_names yes
}
devices {
device {
vendor "(LIO-ORG|PETASAN)"
product "RBD"
path_grouping_policy "multibus"
path_checker "tur"
features "0"
hardware_handler "1 alua"
prio "alua"
failback "immediate"
rr_weight "uniform"
no_path_retry 12
rr_min_io 100
}
}
Is there any other setting should be take care of?
I also notice dmesg info of following:
[ 617.801221] scsi 7:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.802071] scsi 8:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.838207] sd 7:0:0:0: Attached scsi generic sg1 type 0
[ 617.840390] sd 7:0:0:0: [sdb] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.840491] sd 7:0:0:0: [sdb] Write Protect is off
[ 617.840497] sd 7:0:0:0: [sdb] Mode Sense: 43 00 00 08
[ 617.840722] sd 7:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.851077] sd 7:0:0:0: [sdb] Attached SCSI disk
[ 617.857993] sd 8:0:0:0: Attached scsi generic sg2 type 0
[ 617.859795] sd 8:0:0:0: [sdc] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.859914] sd 8:0:0:0: [sdc] Write Protect is off
[ 617.859917] sd 8:0:0:0: [sdc] Mode Sense: 43 00 00 08
[ 617.860114] sd 8:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.871043] sd 8:0:0:0: [sdc] Attached SCSI disk
[ 618.546081] device-mapper: multipath round-robin: version 1.2.0 loaded
[ 618.554183] alua: device handler registered
[ 618.554504] sd 8:0:0:0: alua: supports implicit TPGS
[ 618.554511] sd 8:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 2
[ 618.554516] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.570041] sd 8:0:0:0: alua: transition timeout set to 60 seconds
[ 618.570049] sd 8:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.616605] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.616642] sd 7:0:0:0: alua: supports implicit TPGS
[ 618.616650] sd 7:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 1
[ 618.616654] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.630160] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.630286] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
Is there anything wrong?
Last edited on April 18, 2018, 9:09 am by shadowlin · #2
admin
2,930 Posts
April 18, 2018, 11:59 amQuote from admin on April 18, 2018, 11:59 amI would not worry about the warning too much. It seems the mpio package you use is older than the kernel on the client. Try to find a newer package if available. Also try setting
rr_min_io 100 ->
rr_min_io 1
I would not worry about the warning too much. It seems the mpio package you use is older than the kernel on the client. Try to find a newer package if available. Also try setting
rr_min_io 100 ->
rr_min_io 1
shadowlin
67 Posts
April 18, 2018, 2:37 pmQuote from shadowlin on April 18, 2018, 2:37 pmThank you.
I did upgrade the kernel of ubuntu 14.0.4 server to 4.12.
I tried to shutdown on path during writing test without float ip.It took about 2mins to failover to another path.Is this because of the client side setting?
Thank you.
I did upgrade the kernel of ubuntu 14.0.4 server to 4.12.
I tried to shutdown on path during writing test without float ip.It took about 2mins to failover to another path.Is this because of the client side setting?
admin
2,930 Posts
April 18, 2018, 3:41 pmQuote from admin on April 18, 2018, 3:41 pmFor Linux , you can adjust the following values in the open-iscsi configuration:
noop_out_interval
noop_out_timeout
replacement_timeout
The total fail-over time is the total of the above values
you can also over-ride the replacement_timeout via the multi-path config value (if defined):
fast_io_fail_tmo
For Linux , you can adjust the following values in the open-iscsi configuration:
noop_out_interval
noop_out_timeout
replacement_timeout
The total fail-over time is the total of the above values
you can also over-ride the replacement_timeout via the multi-path config value (if defined):
fast_io_fail_tmo
Last edited on April 18, 2018, 3:41 pm by admin · #5
Is there a guide of how to set multipath for petasan iscsi disk in linux?
shadowlin
67 Posts
Quote from shadowlin on April 18, 2018, 6:23 amI tired to follow the guild from lrbd wiki but It doesn't work out.The documentation sector of petasan has only guide for windows.
I tired to follow the guild from lrbd wiki but It doesn't work out.The documentation sector of petasan has only guide for windows.
shadowlin
67 Posts
Quote from shadowlin on April 18, 2018, 7:38 amI figured out the problem was the vendor name.
after using this /etc/multipath.conf setting it works
defaults { user_friendly_names yes } devices { device { vendor "(LIO-ORG|PETASAN)" product "RBD" path_grouping_policy "multibus" path_checker "tur" features "0" hardware_handler "1 alua" prio "alua" failback "immediate" rr_weight "uniform" no_path_retry 12 rr_min_io 100 } }Is there any other setting should be take care of?
I also notice dmesg info of following:
[ 617.801221] scsi 7:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.802071] scsi 8:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.838207] sd 7:0:0:0: Attached scsi generic sg1 type 0
[ 617.840390] sd 7:0:0:0: [sdb] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.840491] sd 7:0:0:0: [sdb] Write Protect is off
[ 617.840497] sd 7:0:0:0: [sdb] Mode Sense: 43 00 00 08
[ 617.840722] sd 7:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.851077] sd 7:0:0:0: [sdb] Attached SCSI disk
[ 617.857993] sd 8:0:0:0: Attached scsi generic sg2 type 0
[ 617.859795] sd 8:0:0:0: [sdc] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.859914] sd 8:0:0:0: [sdc] Write Protect is off
[ 617.859917] sd 8:0:0:0: [sdc] Mode Sense: 43 00 00 08
[ 617.860114] sd 8:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.871043] sd 8:0:0:0: [sdc] Attached SCSI disk
[ 618.546081] device-mapper: multipath round-robin: version 1.2.0 loaded
[ 618.554183] alua: device handler registered
[ 618.554504] sd 8:0:0:0: alua: supports implicit TPGS
[ 618.554511] sd 8:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 2
[ 618.554516] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.570041] sd 8:0:0:0: alua: transition timeout set to 60 seconds
[ 618.570049] sd 8:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.616605] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.616642] sd 7:0:0:0: alua: supports implicit TPGS
[ 618.616650] sd 7:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 1
[ 618.616654] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.630160] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.630286] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
Is there anything wrong?
I figured out the problem was the vendor name.
after using this /etc/multipath.conf setting it works
defaults { user_friendly_names yes } devices { device { vendor "(LIO-ORG|PETASAN)" product "RBD" path_grouping_policy "multibus" path_checker "tur" features "0" hardware_handler "1 alua" prio "alua" failback "immediate" rr_weight "uniform" no_path_retry 12 rr_min_io 100 } }
Is there any other setting should be take care of?
I also notice dmesg info of following:
[ 617.801221] scsi 7:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.802071] scsi 8:0:0:0: Direct-Access PETASAN RBD 4.0 PQ: 0 ANSI: 5
[ 617.838207] sd 7:0:0:0: Attached scsi generic sg1 type 0
[ 617.840390] sd 7:0:0:0: [sdb] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.840491] sd 7:0:0:0: [sdb] Write Protect is off
[ 617.840497] sd 7:0:0:0: [sdb] Mode Sense: 43 00 00 08
[ 617.840722] sd 7:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.851077] sd 7:0:0:0: [sdb] Attached SCSI disk
[ 617.857993] sd 8:0:0:0: Attached scsi generic sg2 type 0
[ 617.859795] sd 8:0:0:0: [sdc] 21474836481 512-byte logical blocks: (11.0 TB/10.0 TiB)
[ 617.859914] sd 8:0:0:0: [sdc] Write Protect is off
[ 617.859917] sd 8:0:0:0: [sdc] Mode Sense: 43 00 00 08
[ 617.860114] sd 8:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 617.871043] sd 8:0:0:0: [sdc] Attached SCSI disk
[ 618.546081] device-mapper: multipath round-robin: version 1.2.0 loaded
[ 618.554183] alua: device handler registered
[ 618.554504] sd 8:0:0:0: alua: supports implicit TPGS
[ 618.554511] sd 8:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 2
[ 618.554516] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.570041] sd 8:0:0:0: alua: transition timeout set to 60 seconds
[ 618.570049] sd 8:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.616605] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.616642] sd 7:0:0:0: alua: supports implicit TPGS
[ 618.616650] sd 7:0:0:0: alua: device naa.6001405fac5cf32d6d63af9949e4a8cb port group 0 rel port 1
[ 618.616654] device-mapper: multipath round-robin: repeat_count > 1 is deprecated, using 1 instead
[ 618.630160] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
[ 618.630286] sd 7:0:0:0: alua: port group 00 state A non-preferred supports TOlUSNA
Is there anything wrong?
admin
2,930 Posts
Quote from admin on April 18, 2018, 11:59 amI would not worry about the warning too much. It seems the mpio package you use is older than the kernel on the client. Try to find a newer package if available. Also try setting
rr_min_io 100 ->
rr_min_io 1
I would not worry about the warning too much. It seems the mpio package you use is older than the kernel on the client. Try to find a newer package if available. Also try setting
rr_min_io 100 ->
rr_min_io 1
shadowlin
67 Posts
Quote from shadowlin on April 18, 2018, 2:37 pmThank you.
I did upgrade the kernel of ubuntu 14.0.4 server to 4.12.
I tried to shutdown on path during writing test without float ip.It took about 2mins to failover to another path.Is this because of the client side setting?
Thank you.
I did upgrade the kernel of ubuntu 14.0.4 server to 4.12.
I tried to shutdown on path during writing test without float ip.It took about 2mins to failover to another path.Is this because of the client side setting?
admin
2,930 Posts
Quote from admin on April 18, 2018, 3:41 pmFor Linux , you can adjust the following values in the open-iscsi configuration:
noop_out_interval
noop_out_timeout
replacement_timeoutThe total fail-over time is the total of the above values
you can also over-ride the replacement_timeout via the multi-path config value (if defined):
fast_io_fail_tmo
For Linux , you can adjust the following values in the open-iscsi configuration:
noop_out_interval
noop_out_timeout
replacement_timeout
The total fail-over time is the total of the above values
you can also over-ride the replacement_timeout via the multi-path config value (if defined):
fast_io_fail_tmo