For SpycerNodes (SNO2U, SNO5U) and SpycerMedia Gatewyas (SMGW).
Shutdown Procedure
This procedure will shutdown all of the SpycerNodes and SpycerMedia Gateways directed from one Node.
1. Log into one of the SNO or SMGW Node from ssh.
C:\Users>ssh root@192.168.100.38
root@192.168.100.38's password:
Last login: Thu Aug 18 13:37:50 2022 from 192.168.100.220
2. Stop the protocol services from all SMGW Nodes. (Skip to step 4 if no SMGW Nodes)
It could take a few minutes for each service to stop.
mmces service stop nfs -a
mmces service stop smb -a
mmces service stop obj -a
3. Stop the Spectrum Scale / GPFS daemons from all SMGW Nodes.
mmshutdown -N cesNodes
4. Unmount all filesystems form all the other Nodes.
mmumount all -a
5. Stop the Spectrume Scale / GPFS ddaemons from all the other Nodes.
mmshutdown -a
6. Power off all the SNO and SMGW Nodes.
mmdsh -N GUI_SERVERS "poweroff"
7. Wait a few minutes and confirm shutdown (most of lights will be off and fans no longer running).
A example of a shutdown session log is a the bottom for reference.
Startup Procedure
1. Power on all the SNO Nodes first.
- You may use the IPMI web interface to remotely power up the SNOs, or
- You may press the little recessed power On/Off button with a paper clip at the rear of each of the two SNO controllers:
- Or you can disconnect the two AC power cords, wait 30 seconds and reconnect them, the system will auto power up.
2. Power on all the SMGW Nodes next.
- You may use the iDRAC web interface to remotely power up the SMGW, or
- You may press the power on/off button at the front of the SMGW
3. Allow about 10 minutes for boot to complete, ssh into one of the Nodes to confirm GPFS states are active.
mmgetstate -aL
[root@smgw-100942 ~]# mmgetstate -aL
Node number Node name Quorum Nodes up Total nodes GPFS state Remarks
-------------------------------------------------------------------------------------
1 sno-101514-A 2 3 4 active quorum node
2 sno-101514-B 2 3 4 active quorum node
3 smgw-100942 2 3 4 active quorum node
4 smgw-190001 2 3 4 active
4. If you have SMGWs, ssh to any SNO or SMGW and check protocol service states.
mmces state show -a
[root@smgw-100942 ~]# mmces state show NODE AUTH BLOCK NETWORK HDFS_NAMENODE AUTH_OBJ NFS OBJ SMB CES ------------- ---------- ---------- --------- --------------- ---------- ---------- ---------- --------- ----- smgw-100942 DISABLED DISABLED HEALTHY DISABLED DISABLED DISABLED DISABLED HEALTHY HEALTHY
Shutdown session log example:
[root@sno-101514-A /]# mmces service stop nfs -a
Service NFS is not enabled
mmces service stop: Command failed. Examine previous error messages to determine cause.
[root@sno-101514-A /]# mmces service stop smb -a
smgw-190001: Redirecting to /bin/systemctl stop gpfs-smb.service
smgw-190001: SMB: service succesfully stopped.
smgw-190001: Redirecting to /bin/systemctl stop gpfs-ctdb.service
smgw-190001: CTDB: service succesfully stopped.
smgw-100942: Redirecting to /bin/systemctl stop gpfs-smb.service
smgw-100942: SMB: service succesfully stopped.
smgw-100942: Redirecting to /bin/systemctl stop gpfs-ctdb.service
smgw-100942: CTDB: service succesfully stopped.
[root@sno-101514-A /]# mmces service stop obj -a
Service OBJ is not enabled
mmces service stop: Command failed. Examine previous error messages to determine cause.
[root@sno-101514-A /]# mmshutdown -N cesNodes
Fri Aug 19 11:31:21 PDT 2022: mmshutdown: Starting force unmount of GPFS file systems
Fri Aug 19 11:32:21 PDT 2022: mmshutdown: Unmount not finished after 60 seconds. Waiting 0 more seconds
Fri Aug 19 11:32:21 PDT 2022: mmshutdown: Unmount not finished after 60 seconds
Fri Aug 19 11:32:21 PDT 2022: mmshutdown: Shutting down GPFS daemons
Fri Aug 19 11:33:31 PDT 2022: mmshutdown: Finished
[root@sno-101514-A /]# mmumount all -a
Fri Aug 19 11:35:39 PDT 2022: mmumount: Unmounting file systems ...
[root@sno-101514-A /]# mmshutdown -a
Fri Aug 19 11:36:07 PDT 2022: mmshutdown: Starting force unmount of GPFS file systems
Fri Aug 19 11:36:12 PDT 2022: mmshutdown: Shutting down GPFS daemons
Fri Aug 19 11:36:23 PDT 2022: mmshutdown: Finished
[root@sno-101514-A /]# mmdsh -N GUI_SERVERS "poweroff"
sno-101514-A: Connection to sno-101514-a closed by remote host.
Additional reference:
https://www.ibm.com/docs/en/spectrum-scale/5.0.5?topic=cluster-shutting-down-spectrum-scale
0 Comments