双VIOS,多个VIOC,VIOC的rootvg采用VSCSI,datavg采用NPIV
多路径采用EMC的powerpath
VIOS2的root用户做cfgmgr -v,命令一直执行不完成,VIOS上面hdiskpower5对应的VIOC有宕机,重新启动了.
ps -ef可以查到下面信息
root 12255474 18939924 120 10:57:54 pts/2 0:44 /etc/methods/cfgpowerdisk -l hdiskpower5
root 18939924 18219256 0 10:57:54 pts/2 0:00 cfgmgr -l powerpath0
root 18219256 22413536 0 10:57:53 pts/2 0:00 /usr/sbin/powermt config
root 22413536 24379538 0 10:57:51 pts/2 0:00 cfgmgr -v
进程12255474没有办法kill掉,上面进程的其它进程可以kill掉,由子进程12255474开始kill,最后kill -9 22413536
IDENTIFIER TIMESTAMP T C RESOURCE_NAME DESCRIPTION
DD40E667 0630112617 T S vhost5 VSCSI Client connection lost with I/O pe
LABEL: VIOS_CLIENT_CONNECT
IDENTIFIER: DD40E667
Date/Time: Fri Jun 30 11:26:08 CST 2017
Sequence Number: 12285
Machine Id: 00FAE74F4C00
Node Id: 7xxW_VIOS2
Class: S
Type: TEMP
WPAR: Global
Resource Name: vhost5
Description
VSCSI Client connection lost with I/O pending
Probable Causes
Client Timed out I/O or moved to another server during partition mobility
Failure Causes
Client Timed out I/O or moved to another server during partition mobility
Recommended Actions
None if the client moved or check for I/O issues in the error logs
Detail Data
ERNUM
1000 0080
ABSTRACT
Connection lost with outstanding commands.May result in DMA errors
AREA
Initiator
BUILD INFO
BLD: 1609 06-10:28:47 k2016_36A0
LOCATION
Filename:target_queue.c Function:trans_event Line:544
DATA
rc = 0xFFFFFFFFFFFFFFD8 format = 0x0000000000000001 debit = 0x0000000000000003 schedule_q = 0x0000000000000000
waiting_rsp = 0x0000000000000000 waiting_cmd = 0x0000000000000000
所有的VIOC,除hdiskpower5对应的VIOC异常宕机重启了,VIOC的errpt只有启动记录。其它VIOC正常。
宕机的VIOC启动完成后,对VIOS2重启了,重启过程中,所有的VIOC全正常。
没有在padmin下操作cfgdev扫硬件,是因为powermt命令在root下执行方便。