disk size on /vol_HDB/shared

[Former Member]

Hi all,

When I check disk usage, I observe that 94% on the /vol_HDB/shared volume.

But this time it is coming from following directories

22G on /vol_HDB/shared/HDB/global/hdb/log/mnt00001/

14G on /vol_HDB/shared/HDB/exe/

When I have a look at /vol_HDB/shared/HDB/global/hdb/log/mnt00001/

I see that there are 4 directories (hdb00001, hdb00002, hdb00003, hdb00004) and in each directory there are some dat files also says DO_NOT_TOUCH_FILES_IN_THIS_DIRECTORY as you can see below in the pic.

So I didn't delete these dat files. Should I delete them to make more free space or what do you suggest to me? Or is it normal to have 94% on /vol_HDB/shared ?

Thanks,

Inanc

[former_member183326 · Accepted solution]

Have you tried ALTER SYSTEM RECLAIM LOG?

[Former Member · Accepted solution]

I hadn't tried it before you warned. Now I executed ALTER SYSTEM RECLAIM LOG and it was successfully executed. However, disk size still remains same. Does it need some time to take effect?

[former_member182967 · Accepted solution]

Hi Inanc,

Log volumes are located in directory /vol_HDB/shared/HDB/global/hdb/log/mnt0000x/ which are not too large.

HANA binaries are located in directory /vol_HDB/shared/HDB/exe/ whose usage are in normal situation.

Please use command du -sh to check the directory usage.

Maybe the space is consumed by HANA backups, can you check?

Regards,

Ning

[Former Member · Accepted solution]

Hi Ning,

Thanks for your reply. As you will see in the pic, "global" directory has too much size which is coming from log volumes. I really wonder how to delete these files.

Inanc

[Former Member · Accepted solution]

Hi guys,

I solved the problem as below.

First I changed log_mode to 'overwrite'

ALTER SYSTEM ALTER CONFIGURATION ('global.ini', 'SYSTEM') SET ('persistence', 'log_mode') = 'overwrite' WITH RECONFIGURE;

then I restarted database and executed

ALTER SYSTEM RECLAIM LOG;

then restart database and change log_mode to 'normal' (which was default for me)

ALTER SYSTEM ALTER CONFIGURATION ('global.ini', 'SYSTEM') SET ('persistence', 'log_mode') = 'normal' WITH RECONFIGURE;

then a final restart to database. Now log volume is 54%.

I referenced this webpage by the way

http://www.stechies.com/log-volume-full-log-mode-legacy-database-doesnt-accept-an/

Thanks,

Inanc

[Former Member]

The first thing you should check is the state of the log files in HANA Studio: Administration -> Volumes tab -> and drill down to the log volumes for the service in question.  Check to make sure that the log files are in a "Free" state.  If they are, you can just run ALTER SYSTEM RECLAIM LOG;  If not, you need to investigate why... but ultimately, what you did will work as well for your situation...

[Former Member]

Hi Jason,

I just checked the HANA Studio, state of the logs are free right now as in the picture. However I am not sure about their states before I solved the issue.

I think I took a look at HANA Studio last night before the issue solved, as far as I remember there were some logs which are not Free. I am not sure whether I checked the same logs or about their reason being not free.

[Former Member]

This has the explanations for each state:  SAP HANA Reference: M_LOG_SEGMENTS

Full description

This view describes each allocated log segment and shows its current state and log position range, which is currently contained in this segment.

The following log segment states exist:

• Formatting - The log segment is being formatted and not yet used.
• Preallocated - The log segment has been preallocated, but never used.
• Writing - The log segment is currently being written.
• Closed - The log segment is closed, not backed up and is still required for restart.
• Truncated - The log segment is not required for restart, but has not been backed up.
• BackedUp - The log segment has been backed up, but is still required for restart.
• Free - The log segment has been backed up, it is not required for restart and can be reused.

[Former Member]

They all sound familiar I think some of them were formatting or truncated.. If I encounter it again, I will check this out.

Thanks