Failure phenomenon:
Article reproduced from EMC Chinese Support forum
· The EMC grab script runs to the following steps to stop responding and even cause the server to reboot
Copying file/sys/./class/fc_host/host6/statistics/fcp_output_megabytes
Copying File/proc/./meminfo
Copying File/proc/./uptime
Copying File/proc/./loadavg
Copying File/proc/driver/./dsa/trace_ctl
Copying File/proc/driver/./dsa/trace
· The following error occurred on the VMware ESX host running EMC grab
[Root@vmwareesx1 emcgrab]#./emcgrab.sh
./emcgrab.sh:/tmp/emcgrab/emcgrab.main:line 90:syntax Error near unexpected token ' elif '
./emcgrab.sh:/tmp/emcgrab/emcgrab.main:line: ' elif [${os} = ' HP-UX ']
./emcgrab.sh:line 35:start:command Not found
./emcgrab.sh:line 36:initialise:command Not found
./emcgrab.sh:line 109:legal_notice:command Not found
./emcgrab.sh:line 156:interact:command Not found
./normalize_clarify.awk ' for reading (No such file or directory)
./emcgrab.sh:line 214:module:command Not found
Collecting Inq-inquiry Information
Running Command inq.linux-no_dots .............. ......... ... ................. ........ Command inq.linux-no_dotsrunning for more than 420 seconds
Attempting to kill process 2442
/root/tmp/emcgrab/tools/inq.sh:line 115:2442 killed eval ${line} >>${script_tmp}/${output}.txt 2 >>${ERR_RPT}
Running Command Inq.linux-no_dots-et ....... ............... ............... ........ Command Inq.linux-no_dots-et running for more than 420 seconds
Attempting to kill process 4647
/root/tmp/emcgrab/tools/inq.sh:line 115:4647 killed eval ${line} >>${script_tmp}/${output}.txt 2 >>${ERR_RPT}
Solution:
If an error occurs while running EMC Grab (emcreports), or if you run to a half card, confirm the following:
· EMC grab tools for the appropriate platform download
· There is enough space left on the mainframe
· Download directly from the host (or upload via WINSCP) and unzip the EMC grab tool. If the Linux/esx host, and from Windows replication grab tools to the ESX host, note that FTP needs to open binary mode.
· Run a file certificate generation tool such as DSA
· If you are a Dell server and have a Dell PERC h700/h800 SCSI Adapter, you need to install the latest firmware and drivers, and older firmware drivers may cause Windows to restart
· If it is an ESX host, run esxcfg-mpath-l to confirm if there is a dead path (dead path), and rescan if available Hba:esxcfg-rescan
If you still have an error or a stuck case after checking the above items, consider unlocking the tool script file, editing and commenting out the wrong commands, and then doing the following steps:
1. In the CMD run window press CTRL + C, the script will stop executing and ask you "Terminate batch job y/n?"
2. Press "N" Emcreports will skip the current command to execute the next command
1. Extract Emcrpts_xxx_xxx.exe to a directory
2. Find and use a text editor such as Notepad to open emcrpt.cmd or emulex.cmd, comment out or remove the wrong steps or commands
3. Save the script file and click Emcrpt.cmd to run it directly
1. Access to .../emcgrab/tools directory
2. Find the corresponding script file xxx.sh, use VI to open and comment out the wrong steps or commands
3. Save the file and execute it again./emcgrab.sh