Unsolved
21 Posts
0
1365
August 20th, 2019 00:00
Load Balancing (Networker And DataDomain)
Hello my problems: jobs keep in waiting and retry thousand time... I Have a DataDomain 9500 for backup in Main site and one jukebox (scalar) in DR Site for cloning, and EMC Networker 9.1.0.6.... for example before we created 5 device in data domain and assigned them as a "Pool" and our jobs completed with some errors or monitoring bugs(like estimated time and size), for better performance now I create one device in DD and assigned it to one pool,for writing backup in DD in main site and cloning to DR site, in this process some jobs retry many times if clonig is running on the pool.( what is best practice in DD device ? is it better using one Device or many device for backup in DD and cloning in same time without problem,waiting and retry?
No Events found!
hamed1987
21 Posts
0
August 20th, 2019 01:00
I defined two devices in Data Domain ( A - B )
A is one device - B is 5 device in DD
for example :
Policy (AD) > Workflow(AD) >>>>> backup On A + after that Clone to DRsite (Scalar)
Policy (Exchange) >Workflow (Exchange) >>>>> backup on A + after that Clone to DRsite(scalar)
.
.
.
the problem and many retries when happened that cloning time , do we have a limit session in DD devices?
Do you prefer use one device or many device in DataDomain devices?
I have some clone base job in different workflow, but they will be waiting like this too.
crazyrov
4 Operator
•
1.3K Posts
0
August 20th, 2019 01:00
I would suggest to create a separate workflow for cloning as you are using a slower device. This way your backup job will be independent of the cloning jobs. Also, its best you upgrade to 18.2.x as the earlier version have bugs related to cloning.
bingo.1
2.4K Posts
0
August 20th, 2019 05:00
I would go for scripted cloning, especially because it is more controllable than anything else. And it will work independently. Also it is good for cloning load distribution.
So I would setup the following scenario:
- create 1 ddboost device per tape drive
- Then write 1 script to clone the backups off each disk volume
- Run the scripts in parallel so that you will get best performance
hamed1987
21 Posts
0
August 20th, 2019 07:00
thanks
I think this is a good solution and logical...
please send for me a sample of true "clone script" format...because i used it always GUI... ( if you can send some useful script)
if you had a document about DD and Networker best practice please send it....
I like to keep EMC Product please help me to solve my problem....
Thnx alot
bingo.1
2.4K Posts
1
August 20th, 2019 08:00
The core of such a script is pretty easy as it consist of only 2 lines. Here a sample framework to only clone fulls:
mminfo -q "level=full,volume=your_dd_volume,savetime>-1days,!incomplete,!suspect,copies=1" -r "ssid" > out_file
nsrclone -s nw_server -J source_snode -d dest_snode -b your_clone_pool -e "ret_date" -S -f out_file
Depending on your scenario, the script may become more or less complex. But these are the 2 lines which matter.
Now run it daily and off you go.