Hello,
I'm configuring a new HP DL380 Gen 9 with the following config:
90 GB RAM, HBA: "p440ar hba mode" Or HBA: "LSI SAS 9305-24i Host Bus Adapter" 10GB NIC, CPU E52620v3
complete config:
1x M3G77A Hewlett Packard Enterprise DL380 Gen9 E5-2620v3 SP8040TV EU Svr
5x 726719-B21 Hewlett Packard Enterprise HP 16GB 2RX4 PC4-2133P-R KIT
1x 727060-B21 HPE FlexFabric 10Gb 2-port 556FLR-SFP+ Adapter
1x SAS 9305-24i Host Bus Adapter
2x Samsung SAMSUNG SSD 850 EVO Basic 250GB
24x MZ-75E1T0B/EU Samsung SAMSUNG SSD 850 EVO Basic 1TB
I tested this with 2 raidz1 groups in stripe like this:
Zpool
raidz1 (12 disks)
raidz1 (12 disks)
The performance is pretty great and I get the real 10GB wire speed, the LSI SAS 9305-24i is a bit faster on small IO than the p440ar with expander.
I use VMware 6.5 with iSCSI LUN's.
The Problem:
When I storage Vmotion a 20GB machine from LUN1 to LUN2 on this new storage server the migration goes with 1 GB per second and I'am happy with that.
but after a while I got these errors (for more see error log):
(1:3:2/2): Task I/O type: 0, Tag: 0x5765527, Tag Type: 1
> (1:3:2/2): ctl_process_done: 345 seconds
> (0:3:2/2): Task I/O type: 0, Tag: 0x2c9cde6, Tag Type: 1
> (0:3:2/2): ctl_process_done: 344 seconds
And the LUN hung for a while
after some investigation I see high kernel usage when I do the following command's:
#iostat 1
I see cpu idle goes to 0
and with
#top -SH
I see the kernel{zio_write_intr_0} for every cpu using a lot of cpu
when I move one 7GB VM it uses between 40 and 60% with no other IO on the storage server.
I tried with and without compression on but it doesn't mather.
I tried Freenas 9.10 stable and the 11 train both the same result.
Can someone point me in the right direction I don't think this is normal behavior?
I'm configuring a new HP DL380 Gen 9 with the following config:
90 GB RAM, HBA: "p440ar hba mode" Or HBA: "LSI SAS 9305-24i Host Bus Adapter" 10GB NIC, CPU E52620v3
complete config:
1x M3G77A Hewlett Packard Enterprise DL380 Gen9 E5-2620v3 SP8040TV EU Svr
5x 726719-B21 Hewlett Packard Enterprise HP 16GB 2RX4 PC4-2133P-R KIT
1x 727060-B21 HPE FlexFabric 10Gb 2-port 556FLR-SFP+ Adapter
1x SAS 9305-24i Host Bus Adapter
2x Samsung SAMSUNG SSD 850 EVO Basic 250GB
24x MZ-75E1T0B/EU Samsung SAMSUNG SSD 850 EVO Basic 1TB
I tested this with 2 raidz1 groups in stripe like this:
Zpool
raidz1 (12 disks)
raidz1 (12 disks)
The performance is pretty great and I get the real 10GB wire speed, the LSI SAS 9305-24i is a bit faster on small IO than the p440ar with expander.
I use VMware 6.5 with iSCSI LUN's.
The Problem:
When I storage Vmotion a 20GB machine from LUN1 to LUN2 on this new storage server the migration goes with 1 GB per second and I'am happy with that.
but after a while I got these errors (for more see error log):
(1:3:2/2): Task I/O type: 0, Tag: 0x5765527, Tag Type: 1
> (1:3:2/2): ctl_process_done: 345 seconds
> (0:3:2/2): Task I/O type: 0, Tag: 0x2c9cde6, Tag Type: 1
> (0:3:2/2): ctl_process_done: 344 seconds
And the LUN hung for a while
after some investigation I see high kernel usage when I do the following command's:
#iostat 1
I see cpu idle goes to 0
and with
#top -SH
I see the kernel{zio_write_intr_0} for every cpu using a lot of cpu
when I move one 7GB VM it uses between 40 and 60% with no other IO on the storage server.
I tried with and without compression on but it doesn't mather.
I tried Freenas 9.10 stable and the 11 train both the same result.
Can someone point me in the right direction I don't think this is normal behavior?
