Can any one tell me daily health check of TSM server?

sprasad
Posts: 24
Joined: Sun Oct 09, 2016 2:52 pm

Can any one tell me daily health check of TSM server?

Postby sprasad » Wed Oct 12, 2016 6:27 pm

Can any one tell me daily health check of TSM server?

San
Posts: 24
Joined: Sun Oct 09, 2016 2:53 pm

Re: Can any one tell me daily health check of TSM server?

Postby San » Wed Oct 12, 2016 6:28 pm

Daily health check :
It is essential to check the following every day,
i) q db , check that DB size < 80%
Frequency: once in a day.
ii) q log, check that PctUtil < 50% (after the DB-Backup the value should tend to
0%!)
Frequency: once in a day.
iii) q stgpool, check that DISKPOOL is not always ~100%
Frequency: once in a day.
iv) Select count(*) from libvolumes where status=’Scratch’, check the number of
available scratch volumes. Make sure enough scratch volumes are available,
minimum requirement depends upon the setup.
Frequency: Before the backup starts.
If the number of scratch volumes are less than the minimum requirement, then we
need to increase the number of scratch volume to do this,
First identify the volumes that are less utilized by using the below command,
select volume_name, pct_utilized, status from volumes where pct_utilized<10 order
by pct_utilized
the above command will list all the volume names with its current percentage
utilization.
To have a count of number of volumes less than 10%, use
select count(volume_name) from volumes where pct_utilized<10
Then check to which stgpool each volume belongs use,
Q volume <volume name>
Then move the particular volume to another volume belonging to the same storage
pool by using ,
Move data <volume to be moved> <stgpool to which it belongs>
Some times the tapes will not have any data stored in them but the status will
not show as scratch, under these circumstances we have to first identify those volumes
which does not contain any data but the status does not show as scratch, for this we
can use the command
select volume_name from libvolumes where status='Private' and
libvolumes.volume_name not in (select volume_name from volumes) and
libvolumes.volume_name not in (select volume_name from volhistory where type in
('REMOTE', 'BACKUPFULL', 'BACKUPINCR', 'DBSNAPSHOT', 'EXPORT'))
This will display the volume names then we have to update the tape library with these
volume names and set the status as scratch, for this we can use the command
update libv <library name> <volume name> status=scr
v) check if any volumes were put to readonly, to check this use
select 'update vol',volume_name,'access=readwrite' from volumes where
access!='READWRITE'
normal return should be:
ANR2034E SELECT: No match found using this criterion.
ANS8001I Return code 11.
If there are some volumes in read only or unavailable mode, then we will see
something like this,
Unnamed [1] VOLUME_NAME Unnamed [3]
---------- ------------------ ----------------
update vol L20597 access=readwrite
update vol L20781 access=readwrite
update vol L22023 access=readwrite
These volumes can be set to read write mode by using the below command. RW:
example,
UPD VOL < L20597 > ACCESS=READWRITE
After updating the volume we need to perform an audit on these volumes to do this
use the command,
Audit vol <volume name> fix=yes.
Before performing an audit make sure that the particular volume is not mounted,
check the q mount command before performing using audit command.
If any particular volume keeps on going to readonly or unavailable mode then the
particular content has to be moved to another volume using,
Move data <source volume> <destination volume> <stg pool name>
Frequency: Once in 3-4 hours.
vi) Check whether all the administrative schedules have completed successfully, use
the following command to check the same.
Q event * begind=-1 type=admin.
Make sure that all admin schedules complete with out errors.
Frequency: once in a day.
vii) Make sure all the drives are online, to check the current status use
Q drive.
All the drives that are listed should be in online state. If some of them are offline are
in polling state it should be set to online using the below command.
Update drive <library name> <drive name> online=yes
This will update the drive and keep it in online state.
Frequency: Once in 3-4 hours.
viii) Make sure all the paths are online. To check the current status use the following
command,
Q path.
This will list all the paths available and their current status.
Update path source name destination name source type=<source type name>
destination type=<destination type name> library=<library name> online=yes.
Frequency: Once in 3-4 hours.
ix) Check for the list of client backups that have failed over the last 24 hours, use the
following command,
Q event * * begind=-1 endd=today ex=yes.
Get the list of backups that have failed over the last 24 hours and act accordingly.
Frequency: Once in a day.
x) To check if the password for any node is locked.
Use, Q Node
From the output check the last 2 columns, namely days since last access and locked?
Ideally the days since last access should be <1 and locked? Should be NO.
If the password for the node is locked, then use
Update node <Node name> <new password> passexp=0.
Then connect to the respective TSM client and open the GUI, it will prompt for a
password and type in the password.
Frequency : once in a day.


Return to “TSM”

Who is online

Users browsing this forum: No registered users and 1 guest