Reboot of Exadata Cell Nodes in Non-Rolling
Reboot of Exadata Cell Nodes in Non-Rolling
High Level Steps:
a) Increase Disk Repair Time in ASM, to avoid disks drop from ASM
b) Execute All commands as root user on cell Node.
c) Check disks status from OS prompt by using “cellcli” utility.
d) shutdown all databases and CRS on database nodes( 8 Compute nodes exadbnode01-08).
e) Reboot all cellnodes by using dcli command
f) Bring up CRS and databases on all 8 compute nodes
g) Verify the status of celldisks and cell services on all 14 Compute Nodes.
Steps in Details
-
Setting DISK_REPAIR_TIME
If we need to offline the ASM disks for more than the default time of 3.6 hours then adjust the parameters at ASM Level on database Node as ASM User by Running below commands.
login to database node “ exadbnode01” as oracle user
| . oraenv +ASM1 Sqlplus / as sysasm ALTER DISKGROUP DATAC1 SET ATTRIBUTE ‘DISK_REPAIR_TIME’=’8.5H’; ALTER DISKGROUP RECOC1 SET ATTRIBUTE ‘DISK_REPAIR_TIME’=’8.5H’; ALTER DISKGROUP DBFS_DG SET ATTRIBUTE ‘DISK_REPAIR_TIME’=’8.5H’; |
2)Login to CELLNODE
Login to the Cell exacellnode01 as root , run below commands:
| cellcli -e list celldisk ; Cellcli –e List physicaldisk ; Cellcli –e List Griddisk cellcli -e List griddisk attributes name,asmmodestatus,asmdeactivationoutcome; cellcli -e LIST GRIDDISK ATTRIBUTES name, asmmodestatus; |
3) Shutdown DBs & CRS
Shutdown databases and CRS running on all 8 compute nodes (exadbnode01-08)
4)list griddisk attributes
Execute the command below and the output should show either asmmodestatus=Unknown or asmmodestatus=UNUSED and asmdeactivationoutcome=Yes for all griddisks. Only then is it safe to proceed with restarting the cell:
| cellcli -e list griddisk attributes name,asmmodestatus,asmdeactivationoutcome |
5) Reboot the Cell Node
Reboot the cell node using the Linux reboot command from database Node using dcli command and wait for them to come online.
From database node as root user, run “dcli –l root –g cell_group “shutdown -F -r now”
| dcli –l root –g cell_group “shutdown -F -r now |
6) Startup CRs and DBs
Bring up CRS and databases on all 8 Compute Nodes.
7)Verify all grid disks
Verify all grid disks are Cell services status online using the following command:
| cellcli -e list griddisk attributes name, asmmodestatus |
8) Verify cell services
dcli -l root –g cell_group “service celld status” , and make sure all services are showing as Running.
| dcli -l root –g cell_group “service celld status” |
See also