Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

Netbackup - Differential Incremental not working

Created: 12 Jun 2013 | 13 comments

I tried to run a differential backup after a full backup (in the same policy) but apparently i get back the same number of files backed up as the full backup schedule.

Here's the scenarios i have tested

Initially, I have created 1 file (named notepad1.txt)  in a folder and I noticed after full backup completed, the archive bit attributed to the file been removed, and also in netbackup console, it indicated the numbers of file backed up was 2. I created another 1 file (named notepad2.txt) with A flagged and thus I ran the Differential-incremental backup; once the backup completed, the number of files backed up changed to 3.

In restoration attempt, i have chosen the latest catalog, i.e differential-incremental and as expected, i have seen 2 files been restored, but i believe by right, i should ONLY get notepad2.txt restored

I'd tried several attempts and sometimes, i noticed the differential-incremental backup works when I changed the systems time to the following date (Master, Media and Client servers) but many times, it did not. In my lastest attempt before posting this was having all the systems date changed to 5 days after full backup and it worked

All the servers mentioned above installed with Netbackup 7.0 (running on Windows 2003 32 bits, except Master server on W2k3 64 bits), I pushed them to 7.5 and still get the same result.

kindly advise

Operating Systems:

Comments 13 CommentsJump to latest comment

Farid Shaikh's picture

Hi

This seems to be due to "Time Overlap" Property in Netbackup

Defination of Time Overlap : Specifies the number of minites to add to the date range for incremental backups when you use date based backups. This value compensates for differences in the speed of the clock between the Netbackup client and Server. The default is 60 minutes. This value is used during incremental backups when you use the archive bit and when you examine the create time on folfers. This comparison is done for archive bit-based backups as well as date based backups

In oder to see the correct result of your backups, either change the time overlap from default 60 minutes to 0 and try running the differrential backup after 5 or 10 mins OR wait for 60 minutes before running the next differential backup

Regards

Farid

yong look's picture

hi Farid

For the first time after i changed the Time Overlap, it worked; I can see the subsequent backup with only 2 files; instead of 3 (see pic attached) however all other subsequent attempts (tested 4 times) still backed up all the files; in this case, i am not too sure if that was due to Time Overlapping now

here's the activity log for both full and differential incremental backups, as you can see, i have waited 11 minutes to trigger ths incremental backup

full backup

6/15/2013 10:34:28 AM - requesting resource yl-DSU
6/15/2013 10:34:28 AM - requesting resource master.NBU_CLIENT.MAXJOBS.sql
6/15/2013 10:34:28 AM - requesting resource master.NBU_POLICY.MAXJOBS.yl-pol
6/15/2013 10:34:28 AM - granted resource master.NBU_CLIENT.MAXJOBS.sql
6/15/2013 10:34:28 AM - granted resource master.NBU_POLICY.MAXJOBS.yl-pol
6/15/2013 10:34:28 AM - granted resource MediaID=@aaaaq;Path=C:\yl-destination;MediaServer=media
6/15/2013 10:34:28 AM - granted resource yl-DSU
6/15/2013 10:34:29 AM - estimated 0 Kbytes needed
6/15/2013 10:34:30 AM - started process bpbrm (2248)
6/15/2013 10:34:36 AM - Info bpbrm(pid=2248) sql is the host to backup data from    
6/15/2013 10:34:37 AM - Info bpbrm(pid=2248) reading file list from client       
6/15/2013 10:34:37 AM - connecting
6/15/2013 10:34:39 AM - Info bpbrm(pid=2248) starting bpbkar32 on client        
6/15/2013 10:34:39 AM - Info bpbkar32(pid=0) Backup started          
6/15/2013 10:34:39 AM - Info bptm(pid=2212) start           
6/15/2013 10:34:39 AM - connected; connect time: 00:00:02
6/15/2013 10:34:40 AM - Info bptm(pid=2212) using 262144 data buffer size       
6/15/2013 10:34:40 AM - Info bptm(pid=2212) setting receive network buffer to 1049600 bytes     
6/15/2013 10:34:40 AM - Info bptm(pid=2212) using 30 data buffers        
6/15/2013 10:34:41 AM - Info bptm(pid=2212) start backup          
6/15/2013 10:34:42 AM - Info bptm(pid=2212) backup child process is pid 1488.1188      
6/15/2013 10:34:42 AM - begin writing
6/15/2013 10:34:43 AM - Info bptm(pid=1488) start           
6/15/2013 10:34:55 AM - Info bptm(pid=2212) waited for full buffer 1 times, delayed 766 times   
6/15/2013 10:34:56 AM - Info bptm(pid=2212) EXITING with status 0 <----------       
6/15/2013 10:34:57 AM - Info bpbrm(pid=2248) validating image for client sql       
6/15/2013 10:35:00 AM - Info bpbkar32(pid=0) done. status: 0: the requested operation was successfully completed   
6/15/2013 10:35:00 AM - end writing; write time: 00:00:18
the requested operation was successfully completed(0)

and here's the incremental

6/15/2013 10:34:28 AM - requesting resource yl-DSU
6/15/2013 10:34:28 AM - requesting resource master.NBU_CLIENT.MAXJOBS.sql
6/15/2013 10:34:28 AM - requesting resource master.NBU_POLICY.MAXJOBS.yl-pol
6/15/2013 10:34:28 AM - granted resource master.NBU_CLIENT.MAXJOBS.sql
6/15/2013 10:34:28 AM - granted resource master.NBU_POLICY.MAXJOBS.yl-pol
6/15/2013 10:34:28 AM - granted resource MediaID=@aaaaq;Path=C:\yl-destination;MediaServer=media
6/15/2013 10:34:28 AM - granted resource yl-DSU
6/15/2013 10:34:29 AM - estimated 0 Kbytes needed
6/15/2013 10:34:30 AM - started process bpbrm (2248)
6/15/2013 10:34:36 AM - Info bpbrm(pid=2248) sql is the host to backup data from    
6/15/2013 10:34:37 AM - Info bpbrm(pid=2248) reading file list from client       
6/15/2013 10:34:37 AM - connecting
6/15/2013 10:34:39 AM - Info bpbrm(pid=2248) starting bpbkar32 on client        
6/15/2013 10:34:39 AM - Info bpbkar32(pid=0) Backup started          
6/15/2013 10:34:39 AM - Info bptm(pid=2212) start           
6/15/2013 10:34:39 AM - connected; connect time: 00:00:02
6/15/2013 10:34:40 AM - Info bptm(pid=2212) using 262144 data buffer size       
6/15/2013 10:34:40 AM - Info bptm(pid=2212) setting receive network buffer to 1049600 bytes     
6/15/2013 10:34:40 AM - Info bptm(pid=2212) using 30 data buffers        
6/15/2013 10:34:41 AM - Info bptm(pid=2212) start backup          
6/15/2013 10:34:42 AM - Info bptm(pid=2212) backup child process is pid 1488.1188      
6/15/2013 10:34:42 AM - begin writing
6/15/2013 10:34:43 AM - Info bptm(pid=1488) start           
6/15/2013 10:34:55 AM - Info bptm(pid=2212) waited for full buffer 1 times, delayed 766 times   
6/15/2013 10:34:56 AM - Info bptm(pid=2212) EXITING with status 0 <----------       
6/15/2013 10:34:57 AM - Info bpbrm(pid=2248) validating image for client sql       
6/15/2013 10:35:00 AM - Info bpbkar32(pid=0) done. status: 0: the requested operation was successfully completed   
6/15/2013 10:35:00 AM - end writing; write time: 00:00:18
the requested operation was successfully completed(0)

kindly advise

incre backup.png
mph999's picture

I've seen a similar issue and am currently speaking with Engineering about it, it is under investigation to provide some explanation.

When I hear more, I'll post back.

Martin

Regards,  Martin
 
Setting Logs in NetBackup:
http://www.symantec.com/docs/TECH75805
 
Farid Shaikh's picture

Hi Yong,

What version of NBU you are looking at. Also both the details you copied seems to be from Full backup, can you please copy the differential backup details as well

Regards

Farid

yong look's picture

Hi Farid,

apologise for the mistake i have made while copy the log. i re-create the policy and backups as i have cleanup the backup image prior, and as expect, the result still show the same with differential incremental after waited 11 minutes to trigger the next backup

full backup

6/18/2013 8:38:32 PM - requesting resource dsu
6/18/2013 8:38:32 PM - requesting resource master.NBU_CLIENT.MAXJOBS.sql
6/18/2013 8:38:32 PM - requesting resource master.NBU_POLICY.MAXJOBS.test
6/18/2013 8:38:33 PM - granted resource master.NBU_CLIENT.MAXJOBS.sql
6/18/2013 8:38:33 PM - granted resource master.NBU_POLICY.MAXJOBS.test
6/18/2013 8:38:33 PM - granted resource MediaID=@aaaae;Path=C:\DSU;MediaServer=media
6/18/2013 8:38:33 PM - granted resource dsu
6/18/2013 8:38:33 PM - estimated 0 Kbytes needed
6/18/2013 8:38:35 PM - started process bpbrm (648)
6/18/2013 8:38:41 PM - connecting
6/18/2013 8:38:45 PM - connected; connect time: 00:00:04
6/18/2013 8:38:52 PM - begin writing
6/18/2013 8:39:14 PM - end writing; write time: 00:00:22
the requested operation was successfully completed(0)

incremental

6/18/2013 8:50:44 PM - requesting resource dsu
6/18/2013 8:50:44 PM - requesting resource master.NBU_CLIENT.MAXJOBS.sql
6/18/2013 8:50:44 PM - requesting resource master.NBU_POLICY.MAXJOBS.test
6/18/2013 8:50:44 PM - granted resource master.NBU_CLIENT.MAXJOBS.sql
6/18/2013 8:50:44 PM - granted resource master.NBU_POLICY.MAXJOBS.test
6/18/2013 8:50:44 PM - granted resource MediaID=@aaaae;Path=C:\DSU;MediaServer=media
6/18/2013 8:50:44 PM - granted resource dsu
6/18/2013 8:50:44 PM - estimated 0 Kbytes needed
6/18/2013 8:50:46 PM - started process bpbrm (1852)
6/18/2013 8:50:52 PM - connecting
6/18/2013 8:50:56 PM - connected; connect time: 00:00:04
6/18/2013 8:51:03 PM - begin writing
6/18/2013 8:51:25 PM - end writing; write time: 00:00:22
the requested operation was successfully completed(0)

all my servers (master, media and client) are using 7.0, tested on 7.5.0.5 and still get the same result

thanks

increm backup.jpg
Farid Shaikh's picture

Thanks Yong. Can you also let me know what is the time overlap value set at the moment

Regards

Farid

yong look's picture

hi Farid

upon your advise, i change the time overlap to 0 (initially it was set to 60)

yl

mph999's picture

In your testing, when you have run the full, wait 5 mins, then create your new file, wait another 5 mins then run the incr.  Does this then only backup the single new file.

martin

Regards,  Martin
 
Setting Logs in NetBackup:
http://www.symantec.com/docs/TECH75805
 
yong look's picture

hi Martin

Ne, it backed up both the original and new files, as you can see in the screenshot, the full backup consists only 1 file (and 1 archive bit) but when i ran the incremental, it take 3 files now (original and new file with archive bit)

only once i have seen the incremental backup show 2 files, that when i check the result, it only backup the new file created after full backup (which i waited 10 mins)

yl

yong look's picture

HI Martin,

pls ignore my previous reply as after read your statement carefully, i realised that i did not wait for another 10 minutes (after creating a new file); once after i created the new files, i kick start the incremental backup immediately; i will try it later and let you know.

thanks

yl

Mark_Solutions's picture

yl

Whay are your image cleanup jobs failing / partiall successful?

Just wondering if you have an issue with your catalog that could be affecting things here as well?

Finally ... is there anything we need to know about the server you are running the tests on? Could Anti Virus or any other process be affecting the access times on the files you are testing?

On the host properties for sql is it set to to the archive bit?

Go to Host Properties - Clients and open up sql. Then go to the Windows Client section - Client Settings and see what it is set to use on the right hand side - here it also tells you how long it waits before resetting the archive bit

Authorised Symantec Consultant

Don't forget to "Mark as Solution" if someones advice has solved your issue - and please bring back the Thumbs Up!!.

yong look's picture

hi Mark,

for sack of testing, i deleted my backup schedule and expire the catalog before the clean up; that why i suppose why it show job failed/ partially successful

all my master, media and client servers runned on VMs using VMware workstation, i don't suppose it was due to anti-virus as the incrementally backup worked intially when I change the date to the following day (only once it worked) and 5 days after the full backup

on the client host, the A bit was marked when i created a new file and the existing (which completed full backup) does not have the A bit marked, and yet, the incremental backup took both the files in backup

pls refer to my screenshot attached for the client setting setups; so far i only change the time overlap as per Farid advise, the rest i remain same; and as always, i waited 11 minutes before kick start the next backup (i.e incremental) for testing

regards

yl

client setting.jpg
Siva_TCS's picture

Hi Yong - If you don't have full backup schedule in that specific policy then it will do the full backup everytime even if you configured incremental backup. So, please create a full backup schedule and then do the full backup and then incrementals. it should be fine