06-07-2016 04:32 PM
Hey everyone
I have taken over administering backup of our corporation. At the moment we have policies that are setup to run a full backup once a month, run cumulative once a week and run differential every night.
full backup during weekend.
cumulative one day a week random days chosen.
differential runs every night except weekends.
What I want to do is clear some of the redundant backups. I don't want to do cumulative if that weekend a full backup has already run, and no differential needs to run if cumulative is running in the same day. Does anyone see anything wrong with doing this?
second question is since we have about 1600 policies I don't want to do it through the GUI and I am trying to figure out how to do this through linux terminal. I know about commands bpplsched and bpplschedrep but I can't figure out how to delete or exclude dates. for example if I a policy has cumulative backups running on wed I want to either delete or exclude differential backups on wed. If every 4th week full backups run I don't want to run cumulative on the 4th week.
Thanks for any help in advance.
Shar
06-07-2016 09:58 PM
Hi,
bpplsched is used to create / delete schedules, bpplscherep is used to modify existing schedules. So you'll need both really.
C:\>bpplsched Walker
SCHED Daily 0 1 604800 1 0 0 0 *NULL* 0 1 0 0 0 0 -1 0 0
SCHEDCALENDAR <<Indicates calendar schedule in use
SCHEDWIN 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <<Backup window (each pair of numbers indicates a start time and duration, its blank now so not set)
SCHEDRES *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 1 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
C:\>bpplschedrep Walker Daily -weekday Monday 2
C:\>bpplsched Walker
SCHED Daily 0 1 604800 1 0 0 0 *NULL* 0 1 0 0 0 0 -1 0 0
SCHEDCALENDAR
SCHEDCALDAYOWEEK 2,2 << I added Monday (2) of week 2 to the schedule.
SCHEDWIN 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SCHEDRES *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 1 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
C:\>bpplschedrep Walker Daily -weekday Monday 2 -1 61200 7200
C:\>bpplsched Walker
SCHED Daily 0 1 604800 1 0 0 0 *NULL* 0 1 0 0 0 0 -1 0 0
SCHEDCALENDAR
SCHEDCALDAYOWEEK 2,2
SCHEDWIN 0 0 61200 7200 0 0 0 0 0 0 0 0 0 0 <<I added a start time of 17:00 (3600 * 17) and duration of 2 hours (3600 * 2)
SCHEDRES *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 1 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
C:\>bpplsched Walker -L
Schedule: Daily
Type: FULL (0)
Calendar sched: Enabled
Included Dates-----------
Monday, Week 2
Excluded Dates----------
No specific exclude dates entered
No exclude days of week entered
Retention Level: 1 (2 weeks)
u-wind/o/d: 0 0
Incr Type: DELTA (0)
Alt Read Host: (none defined)
Max Frag Size: 0 MB
Synthetic: 0
Checksum Change Detection: 0
PFI Recovery: 0
Maximum MPX: 1
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 017:00:00 019:00:00 041:00:00 043:00:00
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 000:00:00 000:00:00
Hope that helps.
06-07-2016 10:28 PM
Your first question:
It really depend what is your company's requirement (RTO & RPO) and best practice from your backup team.
Be it a full, cumulative or incremental, they do not have any conflict among them. I do find it weird that you have cumulative run randonly - can be any day?! What is the purpose behind that? If a full backup is completed on Sunday, there is no real advantage to run cumulative on Monday. And how do you determine when to run cumulative - meaning you have to manually change the schedule time??
If you don't really need for cumulative (consolidate your incremental) backup, just ditch it and use full + incremental, make it easier for you to manage.
Your second question:
1600 policies! Again, how did it come to that? Any chance of consolidating them into less? I would not allow my master server to have more than 200 policies.
Reason being to manage is time-consuming, to wait for all policies to load in admin console can sometimes be nerve-wrecking.
Anyway, Riaan has provided you the command to do the change across policies.
I would recommend Marianne's frequency-based schedule if you have concern about having multiple calendar-based schedules running at the same time (redundant backups). Here is the link:
https://www.veritas.com/community/articles/netbackup-frequency-based-scheduling
06-09-2016 02:16 PM
Hi Riaan watsons
Thank you for the reply and sorry for the late reply.
Below is a policy that we have. Could you show me the command how to remove the CUMULATIVE_WEEKLY to not run on the same week as the FULL_MONTHLY please.
SCHED FULL_MONTHLY 0 1 2332800 8 0 0 0 *NULL* 0 1 0 0 0 0 -1 0 0
SCHEDCALENDAR
SCHEDCALDAYOWEEK 7,2
SCHEDWIN 0 0 0 0 0 0 0 0 0 0 0 0 0 172800
SCHEDRES *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 8 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHED CUMULATIVE_WEEKLY 4 1 604800 4 0 0 0 *NULL* 0 0 0 0 0 0 -1 0 0
SCHEDWIN 0 0 0 0 0 0 0 0 0 28800 0 0 0 0
SCHEDRES *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 4 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHED DIFFERENTIAL_DAILY 1 1 86400 4 0 0 0 *NULL* 0 0 0 0 0 0 -1 0 0
SCHEDWIN 0 0 64800 21600 64800 21600 64800 21600 64800 21600 64800 21600 0 0
SCHEDRES DSSU_differentials *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDPOOL *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
SCHEDRL 4 1 1 1 1 1 1 1 1 1
SCHEDFOE 0 0 0 0 0 0 0 0 0 0
SCHEDSG *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL* *NULL*
As your question Watson
we have a script that creates the policies and sets the time and date they will run. The full backup and cumulative backup are chosen as random date that way we are not trying to backup 500TB in once weekend for full backup.
If I ditch the cumulative is full backup once a month enough to make restores?
the reason for 1600 policies is we are trying to divide the amount of backups as mentioned we are backing up close to 500TB if not more and since the data is sensitive we need to have a copy somewhere else and not just on the datacenter. The number of policies floctuates as the number of shows we do (entertainment business) and how big the show is.
06-09-2016 04:25 PM
Hey guys
never mind Figured out how the exclude the dates to not run backups.
06-09-2016 06:32 PM
Looking back to your first post, I guess it's necessary to keep your cumulative, because you only run full backup once a month. (unless you can have it run weekly)
Not sure if I understand the reason behind "running full or cumulative randomly", if you said you set the policies by script, then running them is actually planned rather than random. "Trying to avoid backing up a 500TB for all" sounds also like a planned action, not random. If that's the case, cumulative is OK as long as there is no too much gap between 2 cumulative backups. e.g Every 5 to 7 days a cumulative takes place.
With what you described it sounds like your policyname directly tells you what projects (shows) and date they're on, so you can't really have more generic policyname for different projects (shows). In that case, if you're not managing them all in a single shot, that would still be manageable. Sorry I probably overlooked that.