-
- Novice
- Posts: 4
- Liked: 5 times
- Joined: Apr 02, 2013 12:24 pm
- Full Name: nimes
- Contact:
DDBOOST and DATADOMAIN best practices
I have opened a case about best practice on using ddboost whith datadomain and support tell us about KB 1956. But some questions remains about the number of repository to create for all the jobs, how to set the max task per repository. I think it's a good thing to create un topic and share our experiances about this new veeam fonctionality.
SYSTEM TEAM
NIMES METROPOLE
NIMES METROPOLE
-
- Novice
- Posts: 5
- Liked: never
- Joined: Jan 07, 2015 9:43 am
- Full Name: Marcus Persson
- Contact:
Re: DDBOOST and DATADOMAIN best practices
I still think they need to fix DDBOOST duplication to match how other backup-software works(optimized dup). dd replication just isnt enough when you need diffrent retentions offsite.
-
- Novice
- Posts: 4
- Liked: 5 times
- Joined: Apr 02, 2013 12:24 pm
- Full Name: nimes
- Contact:
Re: DDBOOST and DATADOMAIN best practices
Sorry, this topic is not understood. It is to discuss about best practice and not limit.tehfiskk wrote:I still think they need to fix DDBOOST duplication to match how other backup-software works(optimized dup). dd replication just isnt enough when you need diffrent retentions offsite.
Regards,
SYSTEM TEAM
NIMES METROPOLE
NIMES METROPOLE
-
- Influencer
- Posts: 22
- Liked: 4 times
- Joined: Sep 02, 2014 2:51 pm
- Full Name: Scott Meilicke
- Contact:
Re: DDBOOST and DATADOMAIN best practices
Regarding max tasks per repository, I think it depends on your environment. I have a DD2500 over 1G, and have set the max to 12 connections. Since I have several jobs, each with a max of 4 tasks, I could have 12 or even more threads hitting the DD. I set my remote DD to just 4, in part because I know I won't have more than four backup copies running at once, and in part because that is the default.
I have setup two mtrees per DD. One is for local backups, the other for remote backups. Since I have two app servers, and they know nothing about each other, the two different repositories ensure I can manage them without worrying about stomping on each others jobs. It just makes it cleaner, and doesn't impact the dedupe rates.
Scott
I have setup two mtrees per DD. One is for local backups, the other for remote backups. Since I have two app servers, and they know nothing about each other, the two different repositories ensure I can manage them without worrying about stomping on each others jobs. It just makes it cleaner, and doesn't impact the dedupe rates.
Scott
-
- Novice
- Posts: 4
- Liked: 5 times
- Joined: Apr 02, 2013 12:24 pm
- Full Name: nimes
- Contact:
Re: DDBOOST and DATADOMAIN best practices
Hi,
I have 150 VM to backup with 20 jobs. I'm migrating from 2 nfs repository to 1 ddboost repository.
Our DD2500 is connected over 4 x 1G LAN link. Our Veeam server is virtual and can use only 1G over the LAN.
So I have added a proxy to access the DD2500 with an other 1G network link.
The most of our jobs run every night. I make an incremental backup every day with a synthetic full per week.
I have set maximum task for my ddboost repository to 10.
Now I have migrated many jobs but I have to let them run more time to know if synthetic full works fine.
I'm also in contact with the support for wel tuning of ddboost and datadomain.
I'll be back soon for news.
Oliver
I have 150 VM to backup with 20 jobs. I'm migrating from 2 nfs repository to 1 ddboost repository.
Our DD2500 is connected over 4 x 1G LAN link. Our Veeam server is virtual and can use only 1G over the LAN.
So I have added a proxy to access the DD2500 with an other 1G network link.
The most of our jobs run every night. I make an incremental backup every day with a synthetic full per week.
I have set maximum task for my ddboost repository to 10.
Now I have migrated many jobs but I have to let them run more time to know if synthetic full works fine.
I'm also in contact with the support for wel tuning of ddboost and datadomain.
I'll be back soon for news.
Oliver
SYSTEM TEAM
NIMES METROPOLE
NIMES METROPOLE
-
- Novice
- Posts: 4
- Liked: 5 times
- Joined: Apr 02, 2013 12:24 pm
- Full Name: nimes
- Contact:
Re: DDBOOST and DATADOMAIN best practices
Hi,
before a synthetic full of one of our server with 2.5 TB of disks took 40 hours over an nfs proxy repository, now it take only 2 hours !!!
Oliver
before a synthetic full of one of our server with 2.5 TB of disks took 40 hours over an nfs proxy repository, now it take only 2 hours !!!
Oliver
SYSTEM TEAM
NIMES METROPOLE
NIMES METROPOLE
Who is online
Users browsing this forum: scrat and 78 guests