Applying BizTalk hotfixes/app updates and impact on environment availability

When you run BizTalk server you normally have plenty of mission critical applications running on it. Those that cannot go offline. I get a lot of questions around how to apply to fixes and impact on availability. There is no straight answer for that, it depends on what you are trying to do. I decided to put together a walkthrough on this topic.

Elements that affect BizTalk availability (that are not disasters :-))

- BizTalk Specific:

o Deploying Application

§ Overwriting Existing Application

if you cannot take existing application offline (doesn’t affect the rest of the environment) then you can deploy side by side versions –check:

§ Deploying new applications

1. If deploying applications to new hosts then no impact to existing ones and accordingly will not affect your environment availability.

2. If deploying applications to existing hosts (hosting other applications) a host restart is required. Accordingly this has to be applied on server by server bases (as in take a single server offline, deploy the application then bring it back up).

o Patching BizTalk:

§ If fix contain *.sql (as in it updates the DB) then the IIS/hosts across all servers has to be stopped state before executing the update. The solution around that is:

1. Failover to DR site.

2. Apply fix on main site. Then bring it online

3. Failover to main site.

4. Upgrade the DR site.

§ If the fix does not contain *.sql runtime on each server needs to be restarted (this can be applied on server by server bases).

The following is extracted from:

If the hotfix includes an SQL script (.sql), you must stop the BizTalk services and SQL Server Agent before you install the hotfix. After the hotfix installation is complete, restart the BizTalk services and the SQL Server Agent. If the hotfix does not include an SQL script (.sql), you must restart the BizTalk Host Instance after the hotfix installation is complete.

- SQL Specific

o When installing a service pack on SQL cluster the following practice can be applied ( )- which applies on SQL cluster (patch is applied on server by servers, by taking each server offline and patch it)

- Infrastructure Specific(Windows/IIS etc..):

o Each service back or hotfix has a list of services that needs to be restarted after the fix is applied. If the fix affects a service (or a server restart is required) used by BizTalk (IIS or others). Then the fix should be applied on server by server bases.

Happy new year :-)

find me on twitter: