Adding new disk to GPFS - Very practical

Before Adding new disk to GPFS system, below preparation and knowledge is important and neccesary:

In case you care you should really add the PVID before adding the disk to GPFS. hdisk names mean nothing on AIX thus your hdisk5 might be hdisk6 on the other node. AIX cares about the PVID.
Step 2.1
chdev -l hdiskX -a pv=yes # on all node but NOT in parallel

1. Compare
   2) I really forgot to mention but it is also very important.
   As long as the hdisk names are all the same you will have no problem adding disks to gpfs with given NSD server names.
   If the hdisk names are not the same you might run into problems.
   The point is: In case you are adding a disk to a GPFS and a NSD server name is given the task of adding the disk is DONE ON THAT NSD node ( AFAIK). In case the hdisk names are the same everything is ok. If not and you force the adding of the disk you are into trouble.
   Best practise:

• During mmcrnsd creation do not give a NSD Server name
• Use the mmchnsd command right afterwards to give the new disk the NSD server list

Now the disk can be added to a given GPFS.

 

2: Below steps is the details:

Step 1 :

Allocate Disks from Storage to both the nodes ( Dual presented ), make sure that all disks are fairly uniform in the way it presented.

Step 2 :

Login to the primary node of GPFS Cluster and run below listed commands,

cfgmgr -vl fcs0
cfgmgr -vl fcs1

Check and confirm newly added disks are in available state in server, make sure that these are dual presented in both server.

Step 3 :

Convert newly added disks to compatible to add in GPFS cluster.
create a file, ex : /tmp/clusterdev , which contains something like below,

hdisk5:::dataAndMetadata::drensd3
Here first field is new disks and last field is the new device that you are goind to create.
Then execute the command listed below,
mmcrnsd -F /tmp/clusterdev

Step 4 :

Verifying the newly added disk, issue below listed command,
lspv
You can see a number of disks attached to the machine, check for the newly converted disks and ensure that its showing new device name assiciated with that.
Ex :
hdisk5 00c2a982a3c1e822 drensd3

Step 5 :

create a file which contais something like below,
drensd3:::dataAndMetadata:::
first field is the device created for new disk.
ex:
echo "drensd3:::dataAndMetadata:::" > /tmp/diskdesc

Step 6 :

Add the disk in to the GPFS filesystem,
mmadddisk hpqc -F /tmp/diskdesc
This will increase the filesystem size of your GPFS filesystem hpqc, please verify with df -g command.

 

Last step:

Re-balancing the data

In some cases you may wish to have GPFS re-balance existing data over the new disks that were added to the file system. Often it is not necessary to manually re-balance the data across the new disks. New data that is added to the file system is correctly striped. Re-striping a large file system requires a large number of insert and delete operations and may affect system performance. Plan to perform this task when system demand is low.

 

1. To re-balance the existing data in the file system use the mmrestripefs command.

> mmrestripefs fs1 -b

2. Use the mmdf command to view the utilization of each disk.