08-18-2011 01:44 AM
Hi,
have some strange behavior since some time.
Without big changes, the Full Weekly backups are really slower than Full Monthly for unexpected reson.
The backup speed is about 15 MB/s for the monthly Fulls but only about 7 MB/s for the weekly backups.
We are working with Master v. 7.0.1 on Windows 2003 R3. And the backup data are on a Windows cluster 2008 SP2 64bit.
No cluster tresspass..
All ideas are welcome
Thanks
08-18-2011 02:11 AM
Hi,
You'll have to investigate your environment. Search for speed/performance on the forum, there are hundreds of post explaining how you can verify the different components in the data path.
Check whether you're not perhaps running both weekly and monthly jobs at the same time.
08-18-2011 02:41 AM
Weekly and monthly jobs are NOT running at the same time...
08-18-2011 02:49 AM
we need to know few more details .....
is monthly & weekly backup happend in off business hours ??
both backup using same tape drive type ??
both backup using same medis server ?
what is system load status at the time of weekly backup & monthly backup ??
08-18-2011 03:47 AM
Hi,
Weekly & Monthly happened during the week-end.
Yes, both are using the same media server and the same drives.
The System load is similar for weekly and monthly backups..
08-18-2011 04:03 AM
in the same Policy &/or using the same sort of settings? I'm thinking along the lines of "Limit jobs per policy"and "Allow multiple data streams" (Policy level restrictions) or "Media multiplexing" (Schedule level restriction) which will determine how many jobs run at the same time - more concurrent jobs = lower transfer rate.
Are there any exclude_lists/include_lists set up that may be different between the two schedules - NB can spend time building its "list" of what needs to be backed up.
Just a couple of thoughts.....
08-18-2011 04:26 AM
Good thoughts o:)
But unfortunately the policies, multiplex & data streams settings are the same. The schedule are also similar, only the frequence & rentention are different.
No exclude/include list at all...
Thanks
08-18-2011 04:40 AM
Are you talking about same amount of data for your monthly & weekly backup?
It's better you provide a screenshot or spreadsheet of those 2 jobs, with the columns of "kilobytes", "number of files" together with the MB/s you already mentioned.
Again, you need to provide more details especially your policy setting (output of bppllist <policy> -L) , than just giving basic info.
If your monthly & weekly are not using the same policy (which is possible), check its policy attribute if "compression" or "encryption" is enabled.
08-18-2011 04:53 AM
Policy Name: NAS_CH9_FS01_F
Options: 0x0
template: FALSE
c_unused1: ?
Names: (none)
Policy Type: MS-Windows (13)
Active: yes
Effective date: 05/20/2009 14:57:09
Client Compress: no
Follow NFS Mnts: no
Backup netwrk drvs:no
Collect TIR info: no
Mult. Data Stream: no
Perform Snapshot Backup: no
Snapshot Method: (none)
Snapshot Method Arguments: (none)
Perform Offhost Backup: no
Backup Copy: 0
Use Data Mover: no
Data Mover Type: 2
Use Alternate Client: no
Alternate Client Name: (none)
Use Virtual Machine: 0
Hyper-V Server Name: (none)
Enable Instant Recovery: no
Policy Priority: 0
Max Jobs/Policy: Unlimited
Disaster Recovery: 0
Collect BMR Info: no
Keyword: (none specified)
Data Classification: -
Residence is Storage Lifecycle Policy: no
Client Encrypt: no
Checkpoint: no
Residence: STU_VTL_CH9_PDP
Volume Pool: Weekly
Server Group: *ANY*
Granular Restore Info: no
Exchange Source attributes: no
Exchange 2010 Preferred Server: (none defined)
Generation: 21
Ignore Client Direct: no
Client/HW/OS/Pri: lsnnasfs01 Windows-x64 Windows2008 0 0 0 0 ?
Include: F:\
Schedule: Full_Weekly
Type: FULL (0)
Frequency: 7 day(s) (604800 seconds)
Maximum MPX: 1
Synthetic: 0
PFI Recovery: 0
Retention Level: 10 (5 weeks)
u-wind/o/d: 0 0
Incr Type: DELTA (0)
Alt Read Host: (none defined)
Max Frag Size: 0 MB
Number Copies: 1
Fail on Error: 0
Residence: (specific storage unit not required)
Volume Pool: (same as policy volume pool)
Server Group: (same as specified for policy)
Residence is Storage Lifecycle Policy: 0
Daily Windows:
Day Open Close W-Open W-Close
Sunday 000:00:00 000:00:00
Monday 000:00:00 000:00:00
Tuesday 000:00:00 000:00:00
Wednesday 000:00:00 000:00:00
Thursday 000:00:00 000:00:00
Friday 000:00:00 000:00:00
Saturday 000:00:00 018:00:00 144:00:00 162:00:00
Schedule: Incremental
Type: INCR (1)
Frequency: 1 day(s) (86400 seconds)
Maximum MPX: 1
Synthetic: 0
PFI Recovery: 0
Retention Level: 10 (5 weeks)
u-wind/o/d: 0 0
Incr Type: DELTA (0)
Alt Read Host: (none defined)
Max Frag Size: 0 MB
Number Copies: 1
Fail on Error: 0
Residence: (specific storage unit not required)
Volume Pool: (same as policy volume pool)
Server Group: (same as specified for policy)
Residence is Storage Lifecycle Policy: 0
Daily Windows:
Day Open Close W-Open W-Close
Sunday 000:00:00 000:00:00
Monday 022:00:00 026:00:00 046:00:00 050:00:00
Tuesday 022:00:00 026:00:00 070:00:00 074:00:00
Wednesday 022:00:00 026:00:00 094:00:00 098:00:00
Thursday 022:00:00 024:00:00 118:00:00 120:00:00
Friday 000:00:00 000:00:00
Saturday 000:00:00 000:00:00
Schedule: Full_Monthly
Type: FULL (0)
Frequency: 35 day(s) (3024000 seconds)
Maximum MPX: 1
Synthetic: 0
PFI Recovery: 0
Retention Level: 8 (1 year)
u-wind/o/d: 0 0
Incr Type: DELTA (0)
Alt Read Host: (none defined)
Max Frag Size: 0 MB
Number Copies: 1
Fail on Error: 0
Residence: (specific storage unit not required)
Volume Pool: Monthly
Server Group: (same as specified for policy)
Residence is Storage Lifecycle Policy: 0
Daily Windows:
Day Open Close W-Open W-Close
Sunday 000:00:00 000:00:00
Monday 000:00:00 000:00:00
Tuesday 000:00:00 000:00:00
Wednesday 000:00:00 000:00:00
Thursday 000:00:00 000:00:00
Friday 000:00:00 000:00:00
Saturday 000:00:00 018:00:00 144:00:00 162:00:00
08-18-2011 05:05 AM
- do you have issues with either the weekly or monthly backups not running?
With a backup window that looks like (if my calculations are correct) Saturday 00:00 to Sunday 18:00 are the backups getting later & later in the window? Just wondering if you are encountering "schedule creep" where the backup doesn't start exactly when the window opens due to the frequency not having expired. (e.g. for a weekly backup we have a frequency of 6 days or less so that it has passed by the time the backup window opens the following Saturday - for monthly we use calendar based)
How do you find that the monthly backup with a frequency of 35 days actually works?
08-18-2011 05:26 AM
THX.
No issue about the backup completion... both finish successfuly..
And the performance issues occurs "only" with the weekly backups...
Weekly taking much more time to complete.. with the same amount of data as monthly (with small difference)..
08-18-2011 05:33 AM
I was just wondering if they seemed to be starting later & later into the backup window that's all as opposed to immediately the window opens (not necessarily directly related to the issue that you are encountering with actual performance)
08-18-2011 05:40 AM
It seems that Netbackup is quite clever because the jobs are starting every 5 weeks at the begin of the backup winodws...
08-18-2011 06:04 AM
Am I going nuts or is there potential to completely skip a month with a 5 week setup?
08-18-2011 06:06 AM
- but then who's to say that the requirement has to be for calendar months?
08-18-2011 06:09 AM
True, its possible, but not "that" probable :p
Then again, this is IT, anything goes.
08-18-2011 06:34 AM
Do you this this 5 weeks schedule could have any impact on the current issue ?
08-18-2011 06:43 AM
I think most people are used to a monthly backup following a more 'natural' monthly schedule e.g. last (or first) day of the month, last (or first) Saturday of month or whatever.
If the 5 week schedule matches your business requirements, then stick with it.
08-18-2011 06:53 AM
Any idea about the issue...? Anyone having the same issue and fixed it ?
This issue sounds really strange for me
08-18-2011 07:04 AM
other than what has already been offered.
The schedules certainly appear to be the same therefore no configuration issues there. Only other thing has already been mentioned i.e. the overall environmental workload when these jobs are taking place - e.g. NetBackup load (master/media), client load, network load etc.
Is it consistently an issue? Has it only just started to happen? Nothing physically a problem - drive needs cleaning & coincidentally this is the one that the monthly uses?
Sorry, grasping at straws now.