Error while recovery DB

[Former Member]

Hi Team ,

While try to restore the a backup to HANA DB  to a Tenat DB  I am gettting  below error

just after I start recovery.

"

RECOVER DATA finished with error:

recovery could not be completed, volume 0, reached log position 0,

Invalid value <2> for BackupSrc_Topology::Version

"

I suspect  it could related to file structure as when I was install the  HANA DB , I had to change the path for

the hana directory  of  /hana/shared  to /hana/shared1 due to space issue.

But now i can't  make changes here.

Could you please help here.

Regards,

Abhinav

[AtulKumarJain]

Hi Abhinav,

Please check the sap note

2188676 - Recovery of a tenant database failed due to missing log area

You may apply the workaround after checking

if you are facing this error scenario and you are certain the log area cannot be accessed due to another technical reason you have to repeat the tenant database recovery using HANA studio and select the “Initialize Log Area” option in the recovery wizard. Thus, the log area will be re-created. Keep in mind that the not backed up log entries will be lost according to the configured recovery point objective (RPO), typically a maximum of 15 minutes.

Best Regards,

Atul

[Former Member]

Thanks Atul

But the   "Initialize Log Area"  by default already selected.

Regards ,

Abhinav

[AtulKumarJain]

Hi Abhinav,

Is there any change in topology before and after backup and recovery.

Are you getting some more information in log?

Br

Atul

[former_member554823]

Hi All,

While NEW HANA 1 DB installation on QUA ( existing server) and copy of prod HANA1 DB on newly built HANA DB

we also faced same issue as -

RECOVER DATA finished with error:
recovery could not be completed, volume 0, reached log position 0,
Invalid value <2> for BackupSrc_Topology::Version
Could not start master name server

Solution : This error itself giving clue to fix this -- check your new HANA DB client version and the source system HAAN DB ( which backup) you want to restore here .. make sure HANA 1 DB client version should be same..

for any discrepencies it will through the error as

Invalid value <2> for BackupSrc_Topology::Version.

to solve this you must uninstall new build HANA DB and install with same HANA DB version ( media) what is used in prod ( source system HANA) DB .. then do recovery it will be suceeeded .

Thank you all .

[OwenLiu]

Thanks for coming to SAP Community for answers. Please post your question as a new question here:

https://answers.sap.com/questions/ask.html Since you're new in asking questions here, check out our tutorial about asking and answering questions (if you haven't already), as it provides tips for preparing questions more effectively, that draw responses from our members. Please note, that your post here won't be answered.

[Former Member]

This message was moderated.

[Former Member]

Dear All ,

Above error is gone & now .getting below .

RECOVER DATA finished with error:

recovery could not be completed, volume 0, reached log position 0,

Cannot open file ""<root>/SRD2016-03-28223005_databackup_0_1" ((mode= R, access= rw-r-----, flags= DIRECT|MUST_EXIST|MULTI_WRITERS|UNALIGNED_SIZE), factory= (root= "/dev/shm/" (access= rw-r-----, flags= <none>, usage= DATA_BACKUP, fs= UNSUPPORTED (tm, config= (async_write_submit_active=auto,async_write_submit_blocks=new,async_read_submit=off,num_submit_queues=1,num_completion_queues=1,size_kernel_io_queue=512,max_parallel_io_requests=64))", rc=22: Invalid argument

[Former Member]

I have already set permission to all files to 777.

yet getting this error

[AtulKumarJain]

Hi Abhinav,

It seems while giving the input to hana stdio in recovery ,path of backup .

Please share that screenshot .Do not select the datafile name only

"SRD2016-03-28223005"

BR

Atul

[Former Member]

Now I am into another error saying

FAIL: process hdbindexserver indexserver-SRD not running

Regards ,

Abhinav

[AtulKumarJain]

Hi Abhinav,

Recovery completed?And your HANA sysetm is not starting?

For starting recovery your system should be down.

BR

Atul

[Former Member]

No this is after recovery failure.

Regards ,

Abhinav

[AtulKumarJain]

Please check indexserver trace and name serevr trace.

BR

Atul

[Former Member]

Hi  Atul ,

I later realized that its issue with the DB version , so i updated the DB to SP10 , but post that

I have getting below error.

RECOVER DATA finished with error:

recovery could not be completed, volume 0, reached log position 0,

Cannot open file ""<root>/GWD2016-04-01213209_databackup_0_1" ((closed, mode= R, access= rw-r-----, flags= DIRECT|MUST_EXIST|MULTI_WRITERS|UNALIGNED_SIZE), factory= (root= "/dev/shm/" (access= rw-r-----, flags= AUTOCREATE_PATH|DISKFULL_ERROR, usage= DATA_BACKUP, fs= UNSUPPORTED (tmpfs), config= (async_write_submit_active=auto,async_write_submit_blocks=new,async_read_submit=off,num_submit_queues=1,num_completion_queues=1,size_kernel_io_queue=512,max_parallel_io_requests=64,min_submit_batch_size=16,max_submit_batch_size=64))", rc=22: Invalid argument

I am looking forward to clear some space from my system , but snot sure which directory to be cleaned

[Former Member]

This is resolved as this turned out to be a issue with the removing un-necessary - files from the backup directory.

Regards,

Abhinav

[former_member182967]

Hello Abhinav,

Did you try to recover a tenant DB from source backup of a non-MDC environment?

If this is the case, currently is not supported.

Regards,

Ning

[Former Member]

Hi Ning ,

Both Source & Target & MDC environment only.

Regards ,

Abhinav

[AtulKumarJain]

Hi Abhinav,

As you said you have changed the dir structure.

So the backup which you are restoring have taken before the dir structure change?

If yes then you have to correct the dir structure and do the recovery.

BR

Atul

[Former Member]

Hi Arul ,

Post this I have changed re-installed the system & recovered to old directory structure only.

Still same error.

Regards ,

Abhinav

[Former Member]

Hi Abhinav,

Please describe the scenario in more detail

Are you recovering a Multitenant backup to another multitenant backup? Is the topology same on source and destination?

Can you share more entries from the Indexserver trace log for the tenant being recovered?

Sunil