Post by tosis on Jul 22, 2013 17:22:46 GMT 7
Hi there All,
(Cross-posting this to forum.thecus.com)
My N5200B Pro has been running FW 2.01.10.17 beta since April with no issues, but I’ve this weekend run into issues in that it’s started randomly dropping a random RAID disk after I recently enabled 9K jumbo frames. Hoping someone could comment from experience.
Questions:
1. Could the disk dropping described below be related to insufficient resources – perhaps RAM? Would a RAM upgrade solve this (note I’m happy with current performance of the device)
2. If a RAM upgrade won’t solve it, reducing the jumbo frame size for the N5200B to 4K (or disabling altogether) might be the only option – thoughts?
3. What is the consensus in the community regarding the stability of FW 3.00.04.05 beta when used with iSCSI? How stable is it with 9K jumbo frames enabled?
I’m still trawling through the forums for answers but hopefully someone would point me in the right direction…
I’ve a home network consisting of a Windows Home Server 2011 (WHS) front-end box backed by a Thecus N5200B Pro and a N5550. The server had a dual-port Intel Pro1000 MT card and both ports from each device are hooked to my Cisco SAN switch which is split into two VLANs. I’ve also enabled 9000byte jumbo frames just recently with the introduction of the N5550 to the SAN.
The N5200B Pro contains 5x WD10EACS-00ZJB0 1TB drives in a RAID 5 configuration, 64K strip size, and using ZFS. I’ve then allocated 99% of the capacity to a single iSCSI target (~3.7TB) presented to the WHS box.
Finally, all storage attached to this WHS box is aggregated and presented as a single volume by DriveBender (see http://www.division-m.com).
I’ve noticed over the past weekend, since introducing jumbo frames to the N2500B Pro configuration, that it’s started randomly dropping RAID drives when under load (mostly recently it took 30 minutes during a mass file transfer from the iSCSI target). The “chosen” drive will drop, it’s LEDs (particularly the blue power LED) blink off for about 3-5 seconds, the N5200 freaks and the alarm buzzer sounds, and then the drive is rediscovered. This then leads to a RAID rebuild that lasts circa 13 hours. The first time this occurred I also noted an entry in the log that indicated swap disk damage had occurred shortly before the drive was dropped.
(Cross-posting this to forum.thecus.com)
My N5200B Pro has been running FW 2.01.10.17 beta since April with no issues, but I’ve this weekend run into issues in that it’s started randomly dropping a random RAID disk after I recently enabled 9K jumbo frames. Hoping someone could comment from experience.
Questions:
1. Could the disk dropping described below be related to insufficient resources – perhaps RAM? Would a RAM upgrade solve this (note I’m happy with current performance of the device)
2. If a RAM upgrade won’t solve it, reducing the jumbo frame size for the N5200B to 4K (or disabling altogether) might be the only option – thoughts?
3. What is the consensus in the community regarding the stability of FW 3.00.04.05 beta when used with iSCSI? How stable is it with 9K jumbo frames enabled?
I’m still trawling through the forums for answers but hopefully someone would point me in the right direction…
I’ve a home network consisting of a Windows Home Server 2011 (WHS) front-end box backed by a Thecus N5200B Pro and a N5550. The server had a dual-port Intel Pro1000 MT card and both ports from each device are hooked to my Cisco SAN switch which is split into two VLANs. I’ve also enabled 9000byte jumbo frames just recently with the introduction of the N5550 to the SAN.
The N5200B Pro contains 5x WD10EACS-00ZJB0 1TB drives in a RAID 5 configuration, 64K strip size, and using ZFS. I’ve then allocated 99% of the capacity to a single iSCSI target (~3.7TB) presented to the WHS box.
Finally, all storage attached to this WHS box is aggregated and presented as a single volume by DriveBender (see http://www.division-m.com).
I’ve noticed over the past weekend, since introducing jumbo frames to the N2500B Pro configuration, that it’s started randomly dropping RAID drives when under load (mostly recently it took 30 minutes during a mass file transfer from the iSCSI target). The “chosen” drive will drop, it’s LEDs (particularly the blue power LED) blink off for about 3-5 seconds, the N5200 freaks and the alarm buzzer sounds, and then the drive is rediscovered. This then leads to a RAID rebuild that lasts circa 13 hours. The first time this occurred I also noted an entry in the log that indicated swap disk damage had occurred shortly before the drive was dropped.