cancel
Showing results for 
Search instead for 
Did you mean: 

Error 99 from an NDMP backup (NBU 6.5.5)

Jay_Son
Level 5

I have 71 of 72 backups complete successfully. 1 of the backups did not and error'd with a 99. The parent process errored with a 99 status. Can I still recover from this image (day/date) from tape?

1 ACCEPTED SOLUTION

Accepted Solutions

Andy_Welburn
Level 6

 successfully"

No you didn't - you said: "I have 71 of 72 backups complete successfully" - to me that means only ONE of the backups didn't complete successfully, not ALL of them.

"I was always told that ANY parent job completing with other than a STATUS 0 or 1 then I CANNOT RESTORE." - WRONG. You can restore from any child job that did complete "successfully" i.e EC=0 or EC=1

"So, are you saying that I cannot restore the other 71 children (streamed) backups?" - NO I'm saying you can restore from these 71 as long as they finished successfully. i.e EC=0 or EC=1

 

EC=0 i.e. job complete fully, all files successfully backed up i.e. the little blue man in the Admin Console GUI

EC=1 i.e. job partially successful, some files not backed up (see job details) i.e. the little yellow man in the Admin Console GUI

From the Troubleshooting Guide (& elsewhere probably):

NetBackup status code: 0
Message: the requested operation was successfully completed Explanation: No problems were detected with the requested operation. Recommended action: No action is needed, unless it was a database backup that was performed through a database extension product (for example, NetBackup for Oracle or NetBackup for SQL Server). In those instances, code 0 means the backup script (that started the backup) ran without error. However, you must check other status as explained in the related NetBackup manual to see if the database was successfully backed up.

NetBackup status code: 1
Message: the requested operation was partially successful
Explanation: A problem was detected that may require corrective action during
the requested operation.
Recommended action: Check the All Log Entries report and also the progress log
(if there is one).
The following are some of the problems that can appear under status code 1:
■A file or a directory path that is more than 1023 characters long. For NetBackup Snapshot Client: the maximum path name length is 1000 characters for snapshot backups, not 1023. When the snapshot is created, a new mount point is added to the beginning of the file path. If the new mount point plus the original file path exceeds 1023 characters, the backup fails with status code 1. The progress log includes the entry “ERR - Skipping long dir path.”
■Cannot open a file. The file may have been locked for some reason.
NetBackup status codes and messages 137
Status codes
■On a UNIX system, NetBackup cannot get the link name of a file.
■On a UNIX system, NetBackup cannot process a sparse file.
■A read error that was encountered in a file.
■File is of an unknown type, or may be hidden.
■On a UNIX system, the lstat system call fails on a file that is eligible to be backed up. This error may be a permission problem.
■On a UNIX system, a file cannot be locked that has mandatory locking enabled

etc etc etc

 

If you want to confirm you can restore - check in the BAR GUI.

***PS***

That's one of the reasons for doing multi-streamed jobs: if one of the streams fails you don't have to do all of the backup again just that failed job.

View solution in original post

8 REPLIES 8

FrSchind
Level 5
Partner

if the status is >1 there will be no valid netbackup image, so you will not be able to restore it with netbackup. the data that was written to the tape until the error ocurred might be recoverable if it's not already overwritten. it depends on the tools the ndmp host has for recovering this, but i think it will be a really hard time doing it. so the possibility for a recover from this image is in my opinion low.

Zailar
Level 5
Partner Accredited

Hi Jay,

What mean 71 of 72 backups complete successfully?. example one NDMP server and 72 file system backup selections than run backup with tick Allow multiple data streams. You could recover just only 71 file system.

Yogesh9881
Level 6
Accredited

I have 71 of 72 backups complete successfully. ---- Can we say 71 backup streams ??

Jay_Son
Level 5

Yes. I meant 71 backup streams completed and 1 failed with a 99. Can I still do a restore with the 71 streams?

Andy_Welburn
Level 6

As long as the backups finished successfully (EC=0) or partially (EC=1) then you should be able to restore.

The parent finished with a 99 because one of the children did, it does not mean that the rest of the backups willl be invalid.

Jay_Son
Level 5

"As long as the backups finished successfully (EC=0) or partially (EC=1) then you should be able to restore." Well, I just told you that my backups did not complete successfully.

I'm confused again. The "parent" job completed with a "status 99" because one of the "child" job completed with a "status 99" (I DO NOT see "EC=0 or EC=1", what are these? and where do I find them?).

I was always told that ANY parent job completing with other than a STATUS 0 or 1 then I CANNOT RESTORE.

So, are you saying that I cannot restore the other 71 children (streamed) backups?

Andy_Welburn
Level 6

 successfully"

No you didn't - you said: "I have 71 of 72 backups complete successfully" - to me that means only ONE of the backups didn't complete successfully, not ALL of them.

"I was always told that ANY parent job completing with other than a STATUS 0 or 1 then I CANNOT RESTORE." - WRONG. You can restore from any child job that did complete "successfully" i.e EC=0 or EC=1

"So, are you saying that I cannot restore the other 71 children (streamed) backups?" - NO I'm saying you can restore from these 71 as long as they finished successfully. i.e EC=0 or EC=1

 

EC=0 i.e. job complete fully, all files successfully backed up i.e. the little blue man in the Admin Console GUI

EC=1 i.e. job partially successful, some files not backed up (see job details) i.e. the little yellow man in the Admin Console GUI

From the Troubleshooting Guide (& elsewhere probably):

NetBackup status code: 0
Message: the requested operation was successfully completed Explanation: No problems were detected with the requested operation. Recommended action: No action is needed, unless it was a database backup that was performed through a database extension product (for example, NetBackup for Oracle or NetBackup for SQL Server). In those instances, code 0 means the backup script (that started the backup) ran without error. However, you must check other status as explained in the related NetBackup manual to see if the database was successfully backed up.

NetBackup status code: 1
Message: the requested operation was partially successful
Explanation: A problem was detected that may require corrective action during
the requested operation.
Recommended action: Check the All Log Entries report and also the progress log
(if there is one).
The following are some of the problems that can appear under status code 1:
■A file or a directory path that is more than 1023 characters long. For NetBackup Snapshot Client: the maximum path name length is 1000 characters for snapshot backups, not 1023. When the snapshot is created, a new mount point is added to the beginning of the file path. If the new mount point plus the original file path exceeds 1023 characters, the backup fails with status code 1. The progress log includes the entry “ERR - Skipping long dir path.”
■Cannot open a file. The file may have been locked for some reason.
NetBackup status codes and messages 137
Status codes
■On a UNIX system, NetBackup cannot get the link name of a file.
■On a UNIX system, NetBackup cannot process a sparse file.
■A read error that was encountered in a file.
■File is of an unknown type, or may be hidden.
■On a UNIX system, the lstat system call fails on a file that is eligible to be backed up. This error may be a permission problem.
■On a UNIX system, a file cannot be locked that has mandatory locking enabled

etc etc etc

 

If you want to confirm you can restore - check in the BAR GUI.

***PS***

That's one of the reasons for doing multi-streamed jobs: if one of the streams fails you don't have to do all of the backup again just that failed job.

Jay_Son
Level 5

Andy, Thank you for the explanation. HAve a great week.