Scheduling Issues

[Ritesh_Arun]

Hi All,

I need your inputs on one of the issue I am currently facing related to scheduling.

Setup:-

Master Server [linux 2.6]

Media Server [Linux 2.6]

NBU Version [7.6.0.4] on both master and media server.

Admin console installed on [Windows 2008 R2 enterprise Machine]

Issue:- I need to understand which time would the backup schedules follow. I am configuring the backups through an admin console installed on a windows machine and the machine is located in EST. However the actual master and media server are located in CST time zone. Now when i configure the backup policy through an admin cosole which is installed on a windows machine located in EST time zone. Which time zone would my policies would follow.

Options 

1) MASTER SERVER time zone which is CST

2) Admin console Server which is EST.

I am bit confused as I see schedules running or triggerening at a random time and hence i am confused. Please let me know your suggestions. 

[Marianne · Accepted solution]

Rather install Java Admin Console on your desktop. 

Java Console allows you to select a Time Zone.

To see how the backup schedules is interpreted by the master, run this command on the master for one of the policies:

bppllist <policy-name> -U.

To predict which schedules will run from now until midnight, run this command on the master:

nbpemreq  -predict_all -date 04/24/2015 23:59:59

nbpemreq  -predict_all -date <mm/dd/yyyy HH:MM:SS>

[Jaime_Vazquez]

Backup are scheduled based on the 'ctime' created for the date. This ctime value is based on GMT time world wide.  When the ctime is converted to local time, it makes use of the local time zone value.

You can see this by running the command "bpdbm -ctime ctime_value" on servers in two different time zones. As an example:

bpdbm -ctime 1424092884

This returns:
1/5/2015 4:21:24 PM   (EST TZ of server)
1/5/2015 3:21:24 PM   (CST TZ of server)

Backup images are tagged by format of "client_name_ctime", where the ctime is calcualted when the job gets scheduled to work (when it gets tagged in the Activity Monitor, if I remember correctly). Backup policies have a start and end time window for the clients listed in the policy. The NBU job scheduler will see a  policy go active based on the start date/time. It will queue up the clients it finds in the active policy at that time.  A backup job will go active when needed resources are seen as available to it. Backup jobs that are still queued up waiting for a resource or signal will fail with a rc=196 when the policy end time is exceeded.