You are on page 1of 6

BLADE replacement

2.1 Preparation for the replacement


Lock the blade at SAF level:
CUDB_21 SC_2_2# cmw-node-lock PL_2_9

Make a copy of the rpm.conf and leave in the rpm.conf just the linux-
control or linux-payload rpm
CUDB_21 SC_2_1# cp /cluster/nodes/9/etc/rpm.conf
/cluster/nodes/9/etc/rpm.conf_FULL
CUDB_21 SC_2_1# grep -ia linux /cluster/nodes/14/etc/rpm.conf_FULL >
/cluster/nodes/14/etc/rpm.conf

Connect and login in expert mode to Active DMX of the Subrack where
the blade to be replaced is

CUDB_21 SC_2_1# ssh -p 2024 expert@10.202.4.150


expert@10.202.4.150's password: expert
Welcome to the DMX CLI
expert connected from 10.202.4.148 using ssh on blade_0_25
expert@blade_0_25 09:29:53>
expert@blade_0_25 09:29:53>show ManagedElement 1 DmxFunctions 1
BladeGroupManagement 1 Group CUDB ShelfSlot Blade 1

ShelfSlot 0-17
Blade 1
operationalState enabled
availabilityStatus noStatus
productNumber "ROJ 208 840/3"
productRevisionState R6A
productName GEP3-HD300
serialNumber "A065094580 "cd
manufacturingDate 2013-12-23Z
vendorName "Ericsson AB"
changeDate 2014-09-03T08:32:50Z
busType ipmi
firstMacAddress 34:07:fb:ed:30:97
consecutiveMacAddresses 12
Power down the node (blade) to be replaced. From the DMX CLI, run the
command below with the specific slot number.
expert@blade_0_25 09:34:39> conf
expert@blade_0_25 09:34:39> set ManagedElement 1 Equipment 1 Shelf 0
Slot 17 Blade 1 administrativeState locked
commit
expert@blade_0_25 09:34:39>exit
expert@blade_0_25 09:29:53>show ManagedElement 1 DmxFunctions 1
BladeGroupManagement 1 Group CUDB ShelfSlot Blade 1
ShelfSlot 0-17
Blade 1
operationalState disabled
availabilityStatus offLine
productNumber "ROJ 208 840/3"
productRevisionState R6A
productName GEP3-HD300
serialNumber "A065094580 "
manufacturingDate 2013-12-23Z
vendorName "Ericsson AB"
changeDate 2014-09-03T08:32:50Z
busType ipmi
firstMacAddress 34:07:fb:ed:30:97
consecutiveMacAddresses 12
Power off the new blade by issuing the following command:
expert@blade_0_25 22:29:39> conf
Entering configuration mode private
[ok][2015-04-20 22:38:09]

[edit]
expert@blade_0_25 22:38:09% set ManagedElement 1 Equipment 1 Shelf 0
Slot 17 Blade 1 administrativeState locked
[ok][2015-04-20 22:38:28]

[edit]
expert@blade_0_25 22:38:28% commit
Commit complete.
[ok][2015-04-20 22:38:35]

[edit]
Power on the blade by issuing the following command:
expert@blade_0_25 09:34:39> set ManagedElement 1 Equipment 1 Shelf 0
Slot 17 Blade 1 administrativeState unlocked
commit

CUDB_21 SC_2_1# cp /cluster/nodes/9/etc/rpm.conf_FULL


/cluster/nodes/9/etc/rpm.conf
CUDB_21 SC_2_1# cmw-node-unlock PL_2_9

CUDB_21 SC_2_2# cluster reboot -n 9

CUDB_21 SC_2_1# cd /opt/ericsson/cudb/OAM/support/bin


CUDB_21 SC_2_1# ./cudbPartTool rebuild -n 9

CUDB_21 SC_2_1# ssh PL_2_9


CUDB_21 SC_2_1# ./cudbPartTool check -n 9
CUDB_21 SC_2_1# cudbManageStore --ds 2 --order restart
CUDB_21 SC_2_1# cudbUnitDataBackupAndRestore -d 2 -n 21

troca de disco

2.1 Preparation for the replacement


Lock the blade at SAF level:
CUDB_21 SC_2_2# cmw-node-lock PL_2_9

Make a copy of the rpm.conf and leave in the rpm.conf just the
linux-control or linux-payload rpm
CUDB_21 SC_2_1# cp /cluster/nodes/9/etc/rpm.conf
/cluster/nodes/9/etc/rpm.conf_FULL
CUDB_21 SC_2_1# grep -ia linux
/cluster/nodes/14/etc/rpm.conf_FULL >
/cluster/nodes/14/etc/rpm.conf

Connect and login in expert mode to Active DMX of the Subrack


where the blade to be replaced is

CUDB_21 SC_2_1# ssh -p 2024 expert@10.202.4.150


expert@10.202.4.150's password: expert
Welcome to the DMX CLI
expert connected from 10.202.4.148 using ssh on blade_0_25
expert@blade_0_25 09:29:53>
expert@blade_0_25 09:29:53>show ManagedElement 1 DmxFunctions 1
BladeGroupManagement 1 Group CUDB ShelfSlot Blade 1

ShelfSlot 0-17
Blade 1
operationalState enabled
availabilityStatus noStatus
productNumber "ROJ 208 840/3"
productRevisionState R6A
productName GEP3-HD300
serialNumber "A065094580 "
manufacturingDate 2013-12-23Z
vendorName "Ericsson AB"
changeDate 2014-09-03T08:32:50Z
busType ipmi
firstMacAddress 34:07:fb:ed:30:97
consecutiveMacAddresses 12
Power down the node (blade) to be replaced. From the DMX CLI,
run the command below with the specific slot number.
expert@blade_0_25 09:34:39> conf
expert@blade_0_25 09:34:39> set ManagedElement 1 Equipment 1
Shelf 0 Slot 17 Blade 1 administrativeState locked
commit
expert@blade_0_25 09:34:39>exit
expert@blade_0_25 09:29:53>show ManagedElement 1 DmxFunctions 1
BladeGroupManagement 1 Group CUDB ShelfSlot Blade 1
ShelfSlot 0-17
Blade 1
operationalState disabled
availabilityStatus offLine
productNumber "ROJ 208 840/3"
productRevisionState R6A
productName GEP3-HD300
serialNumber "A065094580 "
manufacturingDate 2013-12-23Z
vendorName "Ericsson AB"
changeDate 2014-09-03T08:32:50Z
busType ipmi
firstMacAddress 34:07:fb:ed:30:97
consecutiveMacAddresses 12

Power off the new blade by issuing the following command:


expert@blade_0_25 22:29:39> conf
Entering configuration mode private
[ok][2015-04-20 22:38:09]

[edit]
expert@blade_0_25 22:38:09% set ManagedElement 1 Equipment 1
Shelf 0 Slot 17 Blade 1 administrativeState locked
[ok][2015-04-20 22:38:28]
[edit]
expert@blade_0_25 22:38:28% commit
Commit complete.
[ok][2015-04-20 22:38:35]

CUDB_21 SC_2_2# ssh -p 2024 expert@10.202.4.150


expert@10.202.4.150's password:

expert@blade_0_0 09:57:30> conf


Entering configuration mode private
[ok][2018-07-03 09:57:44]

[edit]
expert@blade_0_0 09:57:44% set ManagedElement 1 Equipment 1
Shelf 0 Slot 17 Blade 1 administrativeState unlocked
[ok][2018-07-03 09:58:09]

[edit]
expert@blade_0_0 09:58:09% commit

CUDB_21 SC_2_2# cmw-node-unlock PL_2_9


CUDB_21 SC_2_2# cp /cluster/nodes/9/etc/rpm.conf_FULL
/cluster/nodes/9/etc/rpm.conf
.
.
CUDB_21 SC_2_2# cluster reboot -n 9
Rebooting node 9 (PL_2_9)
Succeeded to execute /sbin/reboot on 9 (PL_2_9)

wait for 10 min

CUDB_21 SC_2_2# ssh PL_2_9


Last login: Tue Jul 3 12:05:49 2018 from sc_2_2
CUDB_21 PL_2_9#
CUDB_21 PL_2_9# df -h
Filesystem Size Used Avail Use% Mounted on
rootfs 2.0G 1.4G 716M 66% /
root 2.0G 1.4G 716M 66% /
tmpfs 12G 696K 12G 1% /dev/shm
shm 12G 696K 12G 1% /dev/shm
192.168.0.100:/.cluster 62G 24G 35G 40% /cluster
/dev/sdb1 147G 875M 139G 1% /local
/dev/sdb2 147G 188M 140G 1% /local2
CUDB_21 PL_2_9# exit
logout
Connection to PL_2_9 closed.
CUDB_21 SC_2_2# ./cudbPartTool check -n 9

CUDB partitioning tool for EBS

-= Cluster filesystem analysis =-

Payload PL_2_9 report:


Everything seems OK.

Done.
CUDB_21 SC_2_2# cudbUnitDataBackupAndRestore -d 2 -n 21

CREATE PART
--------------------------------

creating backup on node 11

cudbManageStore stores to process: ds2 (in dsgroup2).

Starting Backup ...


Launching order Backup for ds2 in dsgroup 2.
Obtaining Mgm Information.
Trying backup on mgmt access 1, wait a moment ...
ndb_mgm 10.22.0.1 2373 -e "START BACKUP 999 WAIT COMPLETED"
..ok
BACKUP-999 renamed in PL_2_9 to
/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-03_12-27
BACKUP-999 renamed in PL_2_10 to
/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-03_12-27

Backup finished successfully for store ds2.


Stores where order backup was successfully completed: ds2.
cudbManageStore command successful.

DIR CREATE PART


--------------------------------

creating backup directory


/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-03_12-27 on
node 21 on blade 10.22.0.9
creating backup directory
/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-03_12-27 on
node 21 on blade 10.22.0.10

COPY PART
--------------------------------

copying backup files from node 11 blade


10.22.0.9:/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-
03_12-27 to node 21 blade
10.22.0.9:/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-
03_12-27
copying backup files from node 11 blade
10.22.0.10:/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-
03_12-27 to node 21 blade
10.22.0.10:/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-
03_12-27
RESTORE PART
--------------------------------

Restoring backup on node 21

cudbManageStore stores to process: ds2 (in dsgroup2).

Launching restore order in CUDB Node 21 to store ds2 in dsgroup


2.
Starting restore in CUDB Node 21 for store ds2, backup path
/local/cudb/mysql/ndbd/backup/BACKUP/BACKUP-2018-07-03_12-27,
sql scripts path
/home/cudb/storageEngine/config/schema/ds/internal/restoreTempSq
l.
Waiting for restore order(s) to be completed in CUDB Node 21 for
stores : ds2.
restore order finished successfully in CUDB Node 21 for store
ds2.
restore order(s) completed in CUDB Node 21 for stores : ds2.
Stores where order restore was successfully completed: ds2.
Closing connections for all blades of DSUnitGroup 2.
cudbManageStore command successful.

Performing cleanup
cudbUnitDataBackupAndRestore successfully ended
CUDB_21 SC_2_2#