Dimebag1
1 Nickel

Powering Down Celerra NS-G8

Jump to solution
Hello,

I need to shut down the whole celerra (cs+dm) but I totally fail to find the right document. The /nas/sbin/nas_halt - command cannot be found. Does it work like this ?

server_cpu server_2 -halt now
server_cpu server_3 -halt now
server_cpu server_4 -halt now
/sbin/init 0


Thanks,
Ingo
Tags (2)
0 Kudos
1 Solution

Accepted Solutions
Rainer_EMC
5 Rhenium

Re: Powering Down Celerra NS-G8

Jump to solution
0 Kudos
12 Replies
Rainer_EMC
5 Rhenium

Re: Powering Down Celerra NS-G8

Jump to solution
0 Kudos
Dimebag1
1 Nickel

Re: Powering Down Celerra NS-G8

Jump to solution
Like I wonder how to find out those things on my own. ;)
0 Kudos
Highlighted
Rainer_EMC
5 Rhenium

Re: Powering Down Celerra NS-G8

Jump to solution
what works for me is:

Opening the Celerra doc CD html page and use the search

Diving into the doc cd to the directory where the PDF version are kept and having a quick look into each one.
sure there's quite a number of them - but if you just read the first couple of pages with concepts you get an idea what the manual is about and if you even need it.

The concepts sections are only a couple of pages and give you a good idea how the system works.

normally then you end up with less than a dozen manuals that you need more often

Note that the PDF manuals dont explain what the Celerra Manager GUI can do, but its help is quite good.
Dimebag1
1 Nickel

Re: Powering Down Celerra NS-G8

Jump to solution
One thing thats bothers me is - how do I get this thing back on ;)

A few reboots do not start the EMC NAS Service - As I expected.
0 Kudos
Dimebag1
1 Nickel

Re: Powering Down Celerra NS-G8

Jump to solution
6 - slot_0 control station ready
- slot_2 powered off
- slot_3 powered off
- slot_4 powered off
0 Kudos
Dimebag1
1 Nickel

Re: Powering Down Celerra NS-G8

Jump to solution
Sorry for so many relys - I would add the Infos to the former post if possible.

I just switched off everything to get the DM out of the powered down state. Thats how it looks like right now:

login as: nasadmin
*** slot_0 control station ready ***
[nasadmin@cs1 ~]$ su
[root@cs1 nasadmin]# mount /nas
[root@cs1 nasadmin]# mount /nbsnas
mount: /dev/nde1 is not a valid block device
[root@cs1 nasadmin]# mount /nas/dos
mount: mount point /nas/dos is a symbolic link to nowhere
[root@cs1 nasadmin]# server_cpu ALL -reboot

Error with file [/nas/server/servers]
File not found or could not be created

Error with file [/nas/sys/nas_errors]
File not found or could not be created

nas_cmd: system execution failed.
Following is the set of recorded exceptions:

******************************** Thread exception *****************************
In thread Child thread 0x9F4CC68 (thread id)
*******************************************************************************
-------------------------------------------------------------------------------
Class / Object Routine Nature of exception Effect
-------------------------------------------------------------------------------
NAS_FILE file_open No such file or directory:
<00000000B7BE7B88> (From FILE) Operating system error. Fail
-------------------------------------------------------------------------------
NAS_FILE file_open
<00000000B7BE7B88> (From FILE) Routine failure. Fail
-------------------------------------------------------------------------------
NAS_FILE create_read_write @2
<00000000B7BE7B88> (From FILE) Routine failure. Rescue
-------------------------------------------------------------------------------
NAS_DB get_file_w_verify_option @5
<00000000B7BE6C70> Routine failure. Fail
-------------------------------------------------------------------------------
NAS_DB get_file @1
<00000000B7BE6C70> Routine failure. Fail
-------------------------------------------------------------------------------
NAS_DB servers_file @2
<00000000B7BE6C70> Routine failure. Fail
-------------------------------------------------------------------------------


nas_cmd: PANIC: caught signal #11 (Segmentation fault) -- Giving up...
[root@cs1 nasadmin]# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/hda3 2063536 896592 1062120 46% /
/dev/hda1 124427 8790 109213 8% /boot
none 1036792 0 1036792 0% /dev/shm
/dev/mapper/emc_vg_pri_ide-emc_lv_home
604736 30380 543636 6% /home
/dev/mapper/emc_vg_pri_ide-emc_lv_celerra_backup
846632 65312 738312 9% /celerra/backup
/dev/mapper/emc_vg_pri_ide-emc_lv_celerra_backendmonitor
7931 1134 6388 16% /celerra/backendmonitor
/dev/hda5 2063504 783788 1174896 41% /nas
[root@cs1 nasadmin]# /nasmcd/getreason
6 - slot_0 control station ready
0 - slot_2 reset
0 - slot_3 reset
0 - slot_4 reset
0 Kudos
nandas
3 Cadmium

Re: Powering Down Celerra NS-G8

Jump to solution
No need to try to reboot the data movers manually or mount the NAS partitions administratively.

If everything else is working fine - you only need to turn on all the circuit breakers in the cabinet - power on the control stations and ensure the blades (data movers) are also having power.

It seems to me that the control volumes are not available to the data movers from where it will boot and load the NAS.

Please ensure that the backend Boot array is also powered on properly - the Fiber Channel switches are ON and the zoning are correct. In other words, ensure that the connectivity and availability to the backend storage from all the data movers are in place.

Please refer to the following doc - which has the power up and power down procedure -

"Celerra Network Server System Operations v5.6.45"

http://powerlink.emc.com/km/live1/en_US/Offering_Technical/Technical_Documentation/300-004-161.pdf

Hope this helps, if not you may open a case with EMC support.
Thanks,
Sandip
0 Kudos
Dimebag1
1 Nickel

Re: Powering Down Celerra NS-G8

Jump to solution
The Dater Movers seem to be held in a reset state and the switch sees no light. Of course everything is turned on and double-checked. We just replaced the fibre channel cabeling with higher graded ones.

I never powered down, I just placed everything to sleep - I need to get everything going again without touching the rack - because we got a srdf remote locating and it takes some time to get there.

The good thing is that is has not happend in an productive enviroment. If we ever need to shut down and restart the whole thing I will keep this scenario in mind and have some support people from EMC on site ;)

The manual says only how to restart the control station or the data movers but my problem is that all those nas_comments fail to work as they seem to be placed on those missing control luns. Its almost impossible for me to nail down the problem to the root cause. The customer PDFs are really thin.

I opened a call yesterday, maybe they got a clue.
0 Kudos
Peter_EMC
3 Silver

Re: Powering Down Celerra NS-G8

Jump to solution
If you are unable to issue a command to reboot the Datamover, powercycle them and monitor the switch ports. If the datamovers does not log into the Fibre Switch, then I would use the old (working) cable.

btw: a /nasmcd/sbin/t2reset reboot -s 2 would do the same for slot_2 ...
0 Kudos