So just look at your good paths and build a grep -v that matches what you need:

This is what we see.   so grep -v "2,  4"



                    n. active,   declustered              state,
 pdisk               total paths     array     free space  remarks
 -----------------   -----------  -----------  ----------  -------
 e1s001                2,  4      DA1          288 GiB     ok
 e1s002                2,  4      DA1          304 GiB     ok
 e1s003                2,  4      DA1          296 GiB     ok
 e1s004                2,  4      DA1          288 GiB     ok
 e1s005                2,  4      DA1          288 GiB     ok
 e1s006                2,  4      DA1          288 GiB     ok



Ed Wahl
OSC


________________________________
From: Wahl, Edward <[email protected]>
Sent: Tuesday, May 10, 2022 4:01 PM
To: gpfsug main discussion list <[email protected]>
Subject: Re: ESS and Disk Paths

Knew I had a better one around here somewhere:

mmlsrecoverygroup <RG-NAME> -L --pdisk | grep "1, 2"


Ed Wahl
OSC


________________________________
From: Wahl, Edward <[email protected]>
Sent: Tuesday, May 10, 2022 3:57 PM
To: gpfsug main discussion list <[email protected]>
Subject: Re: ESS and Disk Paths

There is much you can do with 'lsscsi' and "mmlspdisk" and whatnot.     Or 
something like "mmgetpdisktopology  > /tmp/mmgetpdisktopology.out; topsummary 
/tmp/mmgetpdisktopology.out   > topsummary"

mmlspdisk should all 4 paths with 2 being notenabled.  Look for missing paths?  
This may be the fastest.
" device = 
"//ibmgssio5-hs.ten/dev/sdo(notEnabled),//ibmgssio5-hs.ten/dev/sdfe(notEnabled),//ibmgssio6-hs.ten/dev/sdo,//ibmgssio6-hs.ten/dev/sdfe"
 "


If you are using IBM ESS hardware you can make some guesses and then use the 
Seagate tools to look and see if each SG device has drives.

"for i in `lsscsi -g -v | grep encl | grep 5147 | awk '{print $7 }'`; do echo 
"*** $i ***"; /usr/lpp/mmfs/updates/latest/firmware/enclosure/wbcli $i 
ddump_hid | egrep "FLASH" -B8; echo; done"

We've got 5147-106 drive enclosures so I can check each disk...

"  i=0; while [  $i -lt 106 ]; do echo "+++ $i +++"; 
/usr/lpp/mmfs/updates/latest/firmware/enclosure/wbcli /dev/sg131 
"getdrivestatus $i" | grep "LED     : ON"; sleep 1; let i=i+1; done  "

Lots of options.  Now that said, many of these are slower than the 
gnrhealthcheck for me.  Your milage may vary.


Ed Wahl
OSC

________________________________
From: gpfsug-discuss <[email protected]> on behalf of Kevin D 
Johnson <[email protected]>
Sent: Tuesday, May 10, 2022 3:28 PM
To: gpfsug main discussion list <[email protected]>
Subject: Re: [gpfsug-discuss] ESS and Disk Paths

https://www.ibm.com/docs/en/ess/6.0.2?topic=commands-essfindmissingdisks-command
 Kevin D. Johnson Senior Managing Consultant MBA, MAcc, MS Global Technology 
and Development [email protected] ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ 
‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍ ‍

https://www.ibm.com/docs/en/ess/6.0.2?topic=commands-essfindmissingdisks-command<https://urldefense.com/v3/__https://www.ibm.com/docs/en/ess/6.0.2?topic=commands-essfindmissingdisks-command__;!!KGKeukY!wZw36em9Cso-eeKMybHCsTN2Gvf-QoNn0T6DP0U_jCosVW7mui1AuISTOctCGC20CLKaz6oLf_ik-kw$>

Kevin D. Johnson
Senior Managing Consultant
MBA, MAcc, MS Global Technology and Development

[email protected]
720-349-6199 office
https://ibm.webex.com/meet/kevindjo<https://urldefense.com/v3/__https://ibm.webex.com/meet/kevindjo__;!!KGKeukY!wZw36em9Cso-eeKMybHCsTN2Gvf-QoNn0T6DP0U_jCosVW7mui1AuISTOctCGC20CLKaz6oLZy7tAJY$>
________________________________
From: gpfsug-discuss <[email protected]> on behalf of Peter 
Childs <[email protected]>
Sent: Tuesday, May 10, 2022 3:17 PM
To: gpfsug main discussion list <[email protected]>
Subject: [EXTERNAL] [gpfsug-discuss] ESS and Disk Paths

We have a nearly new ESS5000 (we're still attempting to commission it...)

Anyway, does anyone know a quick way to check if all the paths to the trays are 
working?

We've got 5 trays and one of the Servers can only see one of the 5 trays via 
one path, and its not causing an alert until we run a full "gnrhealthcheck" 
which we only found when attempting to upgrade it.  Looking closer in the gui 
all the disks in that tray are recording 3 paths not 4 but its not causing any 
health alerts.

Don't worry we've got a ticket open and they are attempting to replace the 
cable, I'm just attempting to find a nice quick way to detect the problem and 
know when its resolved rather than waiting 5 minutes while the gnrhealthcheck 
runs.




Peter Childs
We are recruiting please see 
https://www.qmul.ac.uk/jobs/vacancies/items/6949.html<https://urldefense.com/v3/__https://www.qmul.ac.uk/jobs/vacancies/items/6949.html__;!!KGKeukY!wZw36em9Cso-eeKMybHCsTN2Gvf-QoNn0T6DP0U_jCosVW7mui1AuISTOctCGC20CLKaz6oLuG73gf8$>
  for more details
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at gpfsug.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org<https://urldefense.com/v3/__http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org__;!!KGKeukY!wZw36em9Cso-eeKMybHCsTN2Gvf-QoNn0T6DP0U_jCosVW7mui1AuISTOctCGC20CLKaz6oLkDNZmcc$>
_______________________________________________
gpfsug-discuss mailing list
gpfsug-discuss at gpfsug.org
http://gpfsug.org/mailman/listinfo/gpfsug-discuss_gpfsug.org

Reply via email to