Hey bobby,

sorry you're wright. I thought your were talking about the feature in the group ressource (right click on group -> clone).

But you are right, perhaps I can do my clone job with simple networker tools. I do all my other cloning operations with mminfo + nsrclone and because of that I totally forgot about the networker possibility.

I'll try that.

Hey bobby,

I thought about that.

But I can only specify the Cloning Pool, right?

I need to renew the retention policy from 4 weeks to 10 weeks.

Sorry, forgot to mention that in the first place.

Hi Jan,

You shouldn't need to script anything in order to do this as long as you are using a reasonably modern version of NetWorker.  You can setup a scheduled clone job through the NMC (Configuration tab - Clones - right click - create).  This will allow you to schedule the clone for whatever time you want and to define what backup data to include.  You can select to include only the backups from 1 day ago which should solve your problem. 

-Bobby

Hi,

I'm not sure I understood your last message. You can specify what is to be cloned by the criteria: group, client and/or pool, level, time, name.  The destination of the cloning action is a clone pool.  Is that a problem?  You can define a new browse and retention for the new cloned instance of the saveset that is different from the original. 

-Bobby

That is not the solution I'm looking for.

I don't wanna clone save set ID which I dont't need.

It must be possible to just clone the last successful group run. Not save set IDs from the pre-previous run and no save set IDs from the current run.

Networker is driving me nuts, in Netbackup it's just one mouse click.

JayJay9882,

I assume your complaint is that the backup from Friday runs over midnight into Saturday and therefore you have some savesets also getting picked up for cloning when saying last day?

I believe it is possible to put in a time also so you could put in a time which is 1 minute before your group starts and this should just include the savesets from Saturday's backup only.

If this is not the case then what you could do is set up a new group to run only Saturday and link your client to this and set the group to automatically clone. As your backup runs over into Sunday you will clone this and set the retention you require. The backup will finish and then start the clone.

Regards,

Bill Mason

Yeah Bill,

that's exactly my problem.

Second group would be a solution.

For me the best way would be to generate a file with mminfo were all save set IDs are saved. And this IDs I would clone. But as you say there are also save set from the previous save job when I do a mminfo savetime=>24 hours.

It must be possible to do a mminfo which shows me just the save set IDs from the last successful backup job og group "xyz".

JayJay9882,

I think within the cloning configuration side within NMC you can specify a date/time rather than "1 day ago".

within mminfo you can run something like mminfo -t "24 hour ago" -avot and this will report anything within the last 24 hours of the command running. You can use any number between 1 and 24 in the command so you can build the saveset list from this and omit anything from the previous backup providing the saveset time for Friday's backup is outside of 24 hours (which it should be) or reduce as necessary.

Regards,

Bill Mason

Check nsrclone command and its options, you will find there what you are looking for… If you need really complex query combine mminfo with output to file and nsrclone, which can read this file.

Regards

Lukas

Whatever method you do, it is ssid cloning which happens beneath.  Therefore, create mminfo query to build list of ssids that match your wishes and run it against nsrclone command - simple as that. 

Btw, your example above I would modify to be instead of -r ssid,volume to read -r ssid,cloneid | awk '{print $1"/"$2}'. To make sure you pick up the ssids only for certain day run, use sscreate which is between two runs.

Hi,

I agree with Lukas.  It's probably going to be easier to use only nsrclone without mminfo.

An example:

nsrclone -t "2 hours ago" -g "data2" -S || nsrclone -t "14 hours ago" -g "data2" -S

-Bobby

Thank you guys for helping me.

But also that last command is an unclean solution. I will always clone save set IDs I don't wanna clone. I need a solution where I don't put a time (like hours, etc.) in the command, like:

mminfo -q '!incomplete,savetime>='last complete save of the group,group=GROUP_XYZ' -r ssid,volume | grep ".RO" | awk '{print $1}'

I think 'savetime' is just the wrong choice for the query but I don't find any other option.

Now the to be cloned group runs 16 hours. Perhaps it runs 20 hours in a few weeks. Than I have to configure the clone commands again.

It seems to me that I have to build a second group which just runs once a week. And that group I'm going to clone.

Ok, then I try to give you some information about my environment.

I schedule my backup jobs with UC4 (Enterprise Job Scheduling), not with the networker schedule mechanism. So the group starts every day at a different time depending on other jobs within a UC4-process (Snap, Clone, etc.). I cannot work with something like '48 hours ago' or so because the group is running on different times and also with different duration.

I NEED the information about the last successful group (save set IDs). Sure, I can do 'nsrclone  –t 48 hours ago –e 24 hours ago –g ToClone' but that is dirty. There are save set IDs from the previous run included which I don't want.

Is it a possibility to create a file while I'm running my backup were networker dumps all the save set IDs? This file I could use to get my clone job run with.

Hi JayJay9882,

Ok, you'll have to test it, but I think this will work for what you want to do:

mminfo -q "group=Data2" -r nsavetime,savetime,ssid,client,name | sort -r -k 5,5 -k 1,1 | uniq -f 3 | grep -v ssid | awk {'print $3'}

This will give you the list of latest savesets for each unique client and saveset name combination.

-Bobby