IBM 3000 6.0.2, Service Manual

Page: 64 / 88

declustered capacity all vdisk sets defined
recovery group array type total raw free raw free% in the declustered array
-------------- ----------- ---- --------- -------- ----- ------------------------
ess3k_mySN DA1 NVMe 76 TiB 15 TiB 20% vs_ess3k, vs_ess3k_1
vdisk set map memory per server
node class available required required per vdisk set
---------------------------- --------- -------- ----------------------
ess_x86_64_mmvdisk_5_mySN 102 GiB 7080 MiB vs_ess3k (1074 MiB), vs_ess3k_1 (1077 MiB)
[ess3ka ~]#
The customer can use the new space by creating new vdisk sets from the available space. When the
vdisk sets are added to the existing file system (if required), the restripe operation can be run. For
more information, see IBM Spectrum Scale: Administration Guide.
Example: Manually restarting GPFS on the ESS 3000 canisters
You can manually stop and start GPFS to solidify the nodes configuration changes on both canisters.
For configuration changes to take effect, restart GPFS on one canister at a time, and ensure that at least
one of the canisters is always in the active state. It is also important to understand the effect of losing
quorum on the canister during the GPFS recycling.
1. Check the quorum state by issuing the following command:
# mmgetstate -s
A sample output is as follows:
Node number Node name GPFS state
-------------------------------------------
21 ess3ka-ib active
Summary information
---------------------
Number of nodes defined in the cluster: 10
Number of local nodes active in the cluster: 10
Number of remote nodes joined in this cluster: 0
Number of quorum nodes defined in the cluster: 6
Number of quorum nodes active in the cluster: 6
Quorum = 4, Quorum achieved
2. Shut down GPFS on a canister by issuing the following command:
# mmshutdown -N canisterA
3. Start GPFS on a canister by issuing the following command:
# mmstartup -N canisterA
4. Check the state of GPFS on both canisters by issuing the following command:
# mmgetstate -N this ESS 3000 node class
A sample output is as follows:
Node number Node name GPFS state
-------------------------------------------
21 ess3ka-ib arbitrating
22 ess3kb-ib active
5. Issue the following command until GPFS is in the active state on both canisters:
# mmgetstate -N this ESS 3000 node class
A sample output is as follows:
Node number Node name GPFS state
-------------------------------------------
52 IBM Elastic Storage System 3000: Service Guide