In an HADR environment, install online fix pack updates
on members and cluster caching facilities (CFs) one at a time to update
a DB2® pureScale® instance
to a fix pack or special build with a higher code level while the
instance remains available.
About this task
In an HADR environment, you can update one or more members
or cluster caching facilities (CFs) while the remaining members and
CFs continue to process transactions. You must update all members
and CFs in both the primary and standby clusters before
you can commit the changes and update the DB2 pureScale instance.
Procedure
To install an online fix pack update on a DB2 pureScale instance
in a HADR environment:
- Install the online fix pack update on each
of the members in the standby cluster:
- Log on to the member server in the standby cluster as
root user.
- Uncompress the fix pack or special build image in a
directory that is accessible to the instance owner and root user.
To reduce interruption to standby log replay, install the
online fix pack update on all members except the current replay member.
Then install the online fix pack update on the current replay member.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -online -I instance-name -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts.
- Install the online fix pack update on the
secondary CF server in the standby cluster:
- Log on to the secondary CF server in the standby cluster
as root user.
- Uncompress the fix pack or special build image in a
local directory that is accessible to the instance owner and root
user.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -I instance-name -online -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts. The -online parameter is the
default and can be omitted.
- Install the online fix pack update on the
primary CF in the standby cluster:
- Ensure that the secondary CF in the standby cluster is in PEER
state by issuing the following command as the instance owner:
db2instance –list
The secondary
CF must be in PEER state before you update the primary CF.
- Log on to the primary CF server in the standby cluster
as root user.
- Uncompress the fix pack or special build image in a
local directory that is accessible to the instance owner and root
user.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -online -I instance-name -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts.
- Determine whether the online fix pack update
was successful in all members and CFs in the standby cluster by issuing
the following command:
media-dir/installFixPack -check_commit -I instance-name
If
the command output shows any problems, fix them before you continue
with the next step.
- Install the online fix pack update on each
of the members in the primary cluster:
- Log on to the member server in the primary cluster as
root user.
- Uncompress the fix pack or special build image in a
directory that is accessible to the instance owner and root user.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -online -I instance-name -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts.
- Install the online fix pack update on the
secondary CF server in the primary cluster:
- Log on to the secondary CF server in the primary cluster
as root user.
- Uncompress the fix pack or special build image in a
local directory that is accessible to the instance owner and root
user.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -I instance-name -online -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts. The -online parameter is the
default and can be omitted.
- Install the online fix pack update on the
primary CF server in the primary cluster:
- Ensure that the secondary CF in the primary cluster
is in PEER state by issuing the following command as the instance
owner:
db2instance –list
The
secondary CF must be in PEER state before you update the primary CF.
- Log on to the primary CF server in the primary cluster
as root user.
- Uncompress the fix pack or special build image in a
local directory that is accessible to the instance owner and root
user.
- Issue the installFixPack command
as follows:
media-dir/installFixPack -p FP-install-path -online -I instance-name -l log-file-name -t /tmp/trace-file-name
Where FP-install-path is
the directory where you want to install the fix pack and media-dir is
the directory where you uncompressed the fix pack image. FP-install-path must be the same
on all hosts.
- Determine whether the online fix pack update
was successful in all members and CFs in the primary cluster by issuing
the following command:
media-dir/installFixPack -check_commit -I instance-name
If
the command output shows any problems, fix them before you continue
with the next step.
- Commit the online fix pack update in the
standby cluster so that your DB2 pureScale instance
is updated to the new fix pack level by issuing the following command:
media-dir/installFixPack -commit_level -I instance-name -l log-file-name -t /tmp/trace-file-name
- Verify that your instance and databases show the new committed
fix pack level in the standby cluster by issuing the following command:
db2pd -ruStatus
For
example, if you are updating the fix pack level from DB2 Version 10.5
Fix Pack 1 to DB2 Version 10.5 Fix Pack 3, the following text displays
the command output. ROLLING UPDATE STATUS: Disk Value Memory Value
Record Type = INSTANCE
ID = 0
Code Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
Architecture Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
State = [NONE]
Last updated = 2013/11/14:04:38:47
Record Type = MEMBER
ID = 0
Code Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
CECL = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
Architecture Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
CEAL = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
Section Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
State = [NONE]
Last updated = 2013/11/13:03:39:33
coralpib19c.torolab.ibm.com: db2pd -rustatus -localhost ... completed ok
Record Type = MEMBER
ID = 1
Code Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
CECL = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
Architecture Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
CEAL = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
Section Level = V:10 R:5 M:0 F:0 I:0 SB:0 (0x0A05000300000000) V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000)
State = [NONE]
Last updated = 2013/11/13:00:20:02
coralpib19d.torolab.ibm.com: db2pd -rustatus -localhost ... completed ok
Record Type = CF
ID = 128
Code Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
Architecture Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
State = [NONE]
Last updated = 2013/11/13:03:39:34
Record Type = CF
ID = 129
Code Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
Architecture Level = V:10 R:5 M:0 F:3 I:0 SB:0 (0x0A05000300000000) Not Applicable
State = [NONE]
Last updated = 2013/11/13:00:20:03
In the command
output, V:10 R:5 M:0 F:3 specifies the updated fix
pack level namely DB2 Version 10.5 Fix Pack 3.
- Commit the online fix pack update in the
primary cluster so that your DB2 pureScale instance
is updated to the new fix pack level by issuing the following command:
media-dir/installFixPack -commit_level -I instance-name -l log-file-name -t /tmp/trace-file-name
- Verify that your instance and databases show
the new committed fix pack level in the primary cluster by issuing
the following command:
db2pd -ruStatus
The
command output displays text that is similar to the example provided
in step 10.
- If you want to use capabilities specific
to the fix pack, update the system catalog objects in your databases
in the primary cluster:
- Log on as the instance owner.
- For each database in the instance, issue the db2updv105 command
as follows:
db2updv105 -d db-name