DELL erquallogic PS存储控制器故障的分析处理

对应DELL equallogic PS存储系列，如果控制器出现问题，客户第一步要做的事情就是确认控制器的health status，一般来说，primay和seconday controller的health status是不同的，根据这个health status才可以确定下一步对于控制器维修的方向和步骤。

下面是如何检查控制器的health status，串口或者ssh登录控制器

检查主控制器的health status

Ech hs|ecli

检查seconday 控制器的health status

Ipcsh -s "echo hs|ecli"

对于老的版本，如6.0以下，使用ipctunnel命令来检查seconday controller的health status。

详细的报错信息是可以通过类似上面输出中的16进制字符串0x00000000000404来获得详细的错误描述的。对于下面常见的health status，consult L2的可以联系我们(Wechat: StorageExpert)进行进一步的分析处理。

下面是常见的health status，对于不同的health status有不同的处理方法。

• CACHE_SYNCING_CONDITION  -  probable firmware mismatch on CMs. Controller failover will not work. Consult QL/L2. 
• RAID_SET_LOSTBLK_ENTRY_CONDITION  - Lost blocks inserted consult QL/L2
• RAID_SET_FAULTED_CONDITION – disk failure troubleshoot as normal
• B2B_FAILURE_CONDITION  - communication problem between controllers, typical on CM reboot.
• RAID_SPARE_TOO_SMALL – check the hot spare disk sizes (raidtool and diskview –I s)
• SUMO_CHANNEL_CARD_MISSING – identify card and reseat or replace as needed
• SUMO_CHANNEL_CARD_FAILED  - identify card and reseat or replace as needed
• MORE_SPARE_DRIVES_EXPECTED – expected error when disk has failed and hot spare has started rebuilding.
• SPARE_WRONG_TYPE  - only XS or ES arrays can mix drives and they are restricted to RAID 6 accelerated.
• SSD_RAIDSET_HAS_HDD – Cannot mix SAS or SATA disks on RAIDSET
• DRIVE_NOT_APPROVED – non-Dell disks are not supported with firmware 5.0.x and above.  If genuine disk then firmware block is corrupt or the EVPD info is non-Equallogic
• RAID_SET_DOUBLE_FAULTED_CONDITION – Engage QL/L2 and follow MDF procedure
• RAID_LOST_CACHE_CONDITION – Engage QL/L2 (L2 engage L3)
• RAID_ORPHAN_CACHE_CONDITION – Engage QL/L2 (L2 engage L3) – likely MDF, data in cache does not match disks that are online in the RAIDSET.
• BOTH_SUMO_CHANNEL_CARDS_MISSING – consult QL/L2.
• BOTH_SUMO_CHANNEL_CARDS_FAILED  - consult QL/L2
• RAID_SED_UNRESOLVED – You cannot mix SED and non-SED disks in the same enclosure, remove the alien disk or the array will not boot.