cancel
Showing results for 
Search instead for 
Did you mean: 

Netbackup 7.5 restores 2GB data only from fuse based file system

devesh_gupta
Level 3

Hello All,

I am trying to restore data on fuse based file system using netbackup, but as soon as 2GB of data is restored, netbackup issues close call on the restore operation. I am unable to understand why is netbackup issuing close call.

snippet from bptm:


12:04:27.017 [23877] <2> send_structure_data: Index 36 Field m_nKbPerSec Value <49192>
12:04:32.307 [23877] <2> read_data: Total_Kbytes_Read = 1800192
12:04:39.494 [23877] <2> read_data: Total_Kbytes_Read = 2000128
12:04:39.497 [23877] <2> set_job_details: jobData (327)
12:04:39.497 [23877] <2> send_structure_data: Index 36 Field m_nKbPerSec Value <32066>
12:04:42.389 [23877] <2> read_data: sent keepalive signal to bpbrm, reset count to 19
12:04:42.573 [23877] <2> read_data: read short block, bytes = 37376, remainder = 512
12:04:42.662 [23877] <2> read_data: waited for empty buffer 71 times, delayed 716 times
12:04:42.662 [23877] <2> set_job_details: Tfile (327): LOG 1387281882 4 bptm 23877 waited for empty buffer 71 times, delayed 716 times

12:04:42.662 [23877] <2> send_job_file: job ID 327, ftype = 3 msg len = 80, msg = LOG 1387281882 4 bptm 23877 waited for empty buffer 71 times, delayed 716 times

12:04:42.662 [23877] <2> set_job_details: jobData (327)
12:04:42.662 [23877] <2> send_structure_data: Index 36 Field m_nKbPerSec Value <35019>
12:04:42.662 [23877] <4> report_throughput: VBRT 1 23877 4 1 /home/gmishra/Devesh/nbuFix/ostFs/mount/ *NULL* 0 1 0 2097188  2097188 (bptm.c.15011)
12:04:42.662 [23877] <2> read_data: Total Kbytes transferred 2097188
12:04:42.662 [23877] <2> set_job_details: Tfile (327): END_READING 1387281882

12:04:42.662 [23877] <2> send_job_file: job ID 327, ftype = 3 msg len = 23, msg = END_READING 1387281882

12:04:42.662 [23877] <2> bp_sts_close_image: 0x1422860 0 <hydra117_1387258741_C1_F1 0 1> 1 4 0 1 1 0 complete flag (2)(M_READ_COMPLETE ), close type (1)(STS_CLOSEF_IH_COMPLETE ) 11
12:04:42.663 [23877] <2> wait_for_tar: waiting for tar to exit, timeout is 300
12:07:43.371 [24129] <2> SetMaxDataLimit: maximum data size: current=-1 max=-1
 


 

 

snippet from tar log:


12:03:44 (327.001) FRAGNUM = 1, COPYNUM = 1, KBYTES = 10518656, REMAINDER = 0, FIRSTBLK = 0, LASTBLK = 21037311
12:03:44 (327.001) [23877] data socket to bprd is connected (6)
12:03:44 (327.001) [23877] received string: (34 65719 0 0 0 0 0 /root/data/10GB_1/)
12:03:44 (327.001) [23877] received string: (35 65724 1073741824 9 0 9 1073741824 /root/data/10GB_1/test_file0)
12:03:44 (327.001) CKP - 327.001 1387281824 0 0 10 1 0 0 28 /root/data/10GB_1/test_file0
12:09:44 (327.001) [23877] Exit_shm status = 174
12:09:44 (327.001) [23877] parent bptm is gone, child is removing shared memory
12:09:44 (327.001) The following files/folders were not restored:

12:09:44 (327.001) UTF - /root/data/10GB_1/test_file0
12:09:44 (327.001) [23877] received string: (EXIT STATUS 0)
12:09:44 (327.001) INF - TAR EXITING WITH STATUS = 174
12:09:44 (327.001) INF - TAR RESTORED 1 OF 2 FILES SUCCESSFULLY
12:09:44 (327.001) INF - TAR KEPT 0 EXISTING FILES
12:09:44 (327.001) INF - TAR PARTIALLY RESTORED 0 FILES
 


 

6 REPLIES 6

devesh_gupta
Level 3

I have attached the nbu logs in tarball as well.

Marianne
Level 6
Partner    VIP    Accredited Certified

Where was data backed up from?

Possibility that this destination filesystem has a 2GB file size limitation?

Your logs are difficult to read... Please copy log files one-by-one to filename.txt that corresponds to process name.

We need bptm and bpbrm logs on the media server. Please upload copied files (bptm.txt and bpbrm.txt) as File attachments.

What is logging level for client's tar log?
We will need between level 3 and 5 log.

Please copy entire tar log to tar.txt and post as File attachment.

devesh_gupta
Level 3

Hi Marianne,

Sorry, i have corrected the heading of the post.

Actually, scenerio is that I have taken a backup of 10 GB (single file) to a fuse based file system.

After that I have tried to restore this file from the fuse based file system back to ext4 type file system. During restore I am getting error

ext4 does not have 2GB file size limitation.

Verbose level is already set 5 for all the logs.

Also, I have attached the mentioned logs in desired manner.

Mark_Solutions
Level 6
Partner Accredited Certified

This actually looks like a default 5 minute timeout - as if during the restore preparation it hits a timeout

12:04:42.389 [23871] <2> Bpbrm_siginfo_print: 2: delay 0 signo SIGUSR2:12 code 0 pid 23877
12:09:42.419 [23871] <16> bpbrm readline: socket read failed: errno = 62 - Timer expired
 

Check all of the timeout settings on your media server and client and see if increasing them gets this working

devesh_gupta
Level 3

I have checked the .info file created by the netbackup during backup, and there is something wierd about it.

According to it backup done is only 2GB only, whereas i have successfully backed up 10GB data.

snippet of *.info file :

---------------------------------

IS_SLPNAME=
IMO_SIZE=2147516416
IMO_BLOCKSIZE=32768
IMO_STATUS=4

----------------------------------

I think due to this restore only restore file upto this size only. Can anyone tell me how verify this ?

Marianne
Level 6
Partner    VIP    Accredited Certified

Your FUSE filesystem may have limitations...

Please investigate this filesystem's attributes, limitations, etc.

Best to select a fully POSIX compliant filesystem as backup destination.