Discussions
Categories
Groups
Community Home
Categories
INTERNAL ENABLEMENT
POPULAR
PUBLIC CLOUD
PRIVATE CLOUD
Quick Links
MY LINKS
HELPFUL TIPS
Back to website
Home
Web CMS (TeamSite)
Backup Stratedgy
artor
Hi,
I am setting up the backup for TeamSite in the following way.
1. I have two box one is Prod and other is Dev and have a same content store in both Prod and Dev Teamsite and both these boxes are replicating each other through OpenDeploy(which needs to figure out how) whatever the change occurs on Prod box update on the dev box through OD. So if failure is occured on Prod box then the traffic will direct towards Dev box or any loss occurs on Prod box which can be recovered from Dev box or vice versa. Dev box has multi store.
2.Also We have daily tape backup for the whole machine so in case failover occurs on both the machines at the same time so we can recover the machine through that tape.
For the time being requirement to use Teamsite only business hours so in case of few hours down would be acceptable.
Please let me know these scenario would work effectively for backing up the TeamSite or is there any issues involve to use this technique.
TS 6.1 + Solaris 5.8
Regards,
Find more posts tagged with
Comments
Phil
Some things to think about...
1a. For backing store replication, consider network storage (Netapp, SAN, etc) if you have access to it. This gives you many more options for replicating the backing store.
1b. If you are going to redirect traffic to the Dev server you need a configuration management procedure that would allow you to switch between Production-level code and Dev-level code. This procedure would manage the code located in the iw-home directory.
1c. You will need some way to prevent developers from changing code while the Dev box is in production failover mode.
2. When you are running your tape backup on Dev the backing store will be frozen and will prevent any Prod-to-Dev synchronization from occuring during this time. You will probably need to add some logic to retry synchronization when it fails.
artor
Thanks Phill to answer.
1a. For backing store replication, consider network storage (Netapp, SAN, etc) if you have access to it. This gives you many more options for replicating the backing store.
Right now there is no network storage only have two machines.
1b. If you are going to redirect traffic to the Dev server you need a configuration management procedure that would allow you to switch between Production-level code and Dev-level code. This procedure would manage the code located in the iw-home directory.
Actually Dev box has two content store one store maps to the Prod box content store and also check the limitation of config files in iw-home directory or does any body know please let me know
1c. You will need some way to prevent developers from changing code while the Dev box is in production failover mode.
Well for development has serperate content store which I don't think so will effect the development activity in case of failover. if you know some point please let me know
2. When you are running your tape backup on Dev the backing store will be frozen and will prevent any Prod-to-Dev synchronization from occuring during this time. You will probably need to add some logic to retry synchronization when it fails.
Actually for the time being TS is used only in business hour I mean 9 to 5 and backup will be done in the evening time. and I believe iwfreeze is used only for content cannot be updated by users so as for this nobody is using in evening time so no need to freeze it first ...am i right??
Please let me know I am following the right direction or not I would really appreciate your help.
Regards,
A
Phil
Thanks Phill to answer.
1a. For backing store replication, consider network storage (Netapp, SAN, etc) if you have access to it. This gives you many more options for replicating the backing store.
Right now there is no network storage only have two machines.
1b. If you are going to redirect traffic to the Dev server you need a configuration management procedure that would allow you to switch between Production-level code and Dev-level code. This procedure would manage the code located in the iw-home directory.
Actually Dev box has two content store one store maps to the Prod box content store and also check the limitation of config files in iw-home directory or does any body know please let me know
1c. You will need some way to prevent developers from changing code while the Dev box is in production failover mode.
Well for development has serperate content store which I don't think so will effect the development activity in case of failover. if you know some point please let me know
The thing to keep in mind is you have two areas where development is occuring. One is the backing store which you have covered. The other is the iw-home directory where developers could be modifying things like iw.cfg, templating.cfg, etc. You should think about how to deal with the iw-home directory when you are failing over to Dev.
2. When you are running your tape backup on Dev the backing store will be frozen and will prevent any Prod-to-Dev synchronization from occuring during this time. You will probably need to add some logic to retry synchronization when it fails.
Actually for the time being TS is used only in business hour I mean 9 to 5 and backup will be done in the evening time. and I believe iwfreeze is used only for content cannot be updated by users so as for this nobody is using in evening time so no need to freeze it first ...am i right??
You absolutely need to freeze before doing a backup. This ensures that your backing store is in a consistent state during the entire backup.
Please let me know I am following the right direction or not I would really appreciate your help.
Regards,
A
Adam Stoller
You didn't indicate what version of OpenDeploy you were using.
If you are using OD6 - it has a built-in capability for transferring metadata along with content.
If you are using any earlier version of OD - you need to have a DNR script on both sides of the deployment - one to capture all the EAs of the to-be-deployed-files and one to apply those EAs on the have-been-deployed-files.
Frankly, I think investing in a network storage devide as Phil suggested would be well worth investigating. The hardware cost may potentially be far less significant than the development / maintenance costs of what you are proposing.
BTW - I'm assuming you *are* licensed to use MultiStore ...
--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com
artor
Thanks Ghoti to reply.
I am using OD 6.0 and you are right about Hardware cost but actually I am also stuck on my available resources that's why I am looking the soltuion in this way.
Well If you think there is more trouble in this approach please let me know. Thanks both of you for nice input.
Regards,
A