lets imagine an ES1 system, in Exchange environment. There are several components that need patching/updating.
What is the best practice implementing Windows OS updates, in ES1 environment?
When an update needs restart, then how to do that. If everything is redundant, then carefully restarting the servers, ES1 stay operational.
What about Master Service? If I restart it, during a short period there is no job coordination happen. How can ES1 tolerate this?
What is the best practice implementing MS Office (Outlook) updates, in ES1 environment?
3. Customer updates his Exchange environment, Has the Exchange's Cumulative Updates (CU's) any effect to ES1?
4.What is the best practice implementing MS SQL updates, in ES1 environment?
5. Those has an install guide.
If I need to implement a hotfix, is it possible to implement it on servers one-by-one, using the redundancy of the services on several servers, and ES1 will stop only for a short time, when fixing the Master Service?
Answer on all your questions:
It is advised to use the following the steps when rebooting (as example, after implementing a hotfix/patch/update) :
- Run SourceOne Suspend for Backup script (make sure you include DiscoveryManager if this is implemented, by default, this is commented out in the scripts)
- After script has completed (approx. 5 minutes) run a new script which stops all EMC SourceOne Services (make sure you stop Job Scheduler and Address Cache as last) and turns the startup type into 'Disabled or Manual' to prevent these servers from starting again after a reboot.
- After all services are stopped, you can take a back-up of all the necessary components
- Implement Exchange/MS SQL/Windows updates/patches (Including Reboot)
- Test updates and patches
- Change startup type from Sourceone Services to Automatic
- Start SourceOne Services (Job scheduler and Address Cache as first)
- Run SourceOne Resume after backup script.
- Best practice: run an historical archive on a mailbox to test if everything is OK on SourceOne Side