Josh, it sounds like you are getting confused a bit... Let me offer some general info and then some setup info and probably more CPS info than you really want to know: 
A policy using Full and Differential backups for Exchange is not continuous protection. You are however correct in thinking a differential will backup all changes made since the last full backup. This setup you describe is a standard Exchange backup strategy using Fulls and Differentials (or Incrementals) but is not considered Continuous Protection.
To do continuous protection for Exchange you have to have the Continuous Protection server setup. Generally, you want to install the main CPS server on the same server as Backup Exec. From there, you push over the CPA (Continuous Protection Agent) to the Exchange server, making sure the Exchange Protection option is checked. Also, Backup Exec of course has to have it's remote agent pushed to the exchange server as well.
In Backup Exec, make sure all your previous jobs for Exchange protection have been deleted. You cannot mix regular backups and continuous protection backups of exchange due to the log sequence will be messed up by the regular backup jobs and will interfere with the restore side of CPS later.
Once all previous jobs have been removed, create a new backup job as you normally due. Select Exchange as the data to be protected and then set the Exchange job to Full. Now, o n the Exchange node, you will see an option for Continuous Protection. Check that. Now you will see the Recovery Point job option become available. That is the key to CPS. Set the Recovery Point job to something like 4-8 hours. Now you have to assign the job to a B2D folder on an NTFS drive, and also give it a recurring schedule (for the full part of the job) Submit it.
Once the job is submitted, you will see it in Job monitor of Backup Exec. It will always display the schedule for the next Full job to run. The Recovery Point jobs will run as well (you actually just created a job which has two jobs within it, and two schedules as well) on the schedule you gave them (every 4-8 hours) but you will only see them when they are active. In job history you will see recovery point jobs tagged so they can be identified from the Full... such as "<job name>-Recovery Point"
In CPS console, you can go to Job Monitor and you will see the job here as well, but not in job setup. This is because CPS did not create the job, Backup Exec did. You can only modify the job in Backup Exec. In CPS job monitor, you will see it displayed as "<job name>-Backup Exec" which indicates this job is created and controlled by Backup Exec and not CPS.
The reason to use CPS for Exchange is as the transaction logs are created on the exchange server, CPS will copy them over to the BE server. This eliminates the window of failure opportunity where the Exchange server could crash hours before the next backup job ran. With the transaction logs already on the BE server, the Recovery Point jobs will then back them up and trigger the exchange server to check to see if they are ready to be purged (controls transaction log growth).
On the restore side of a CPS job you have a new option: Point in Time Restore. WIth this restore option, you can now just specify that you want t restore up to a specific time, such as 9:42am. You select this feature, then enter the time, and Backup Exec will automatically find all the relevant transaction logs from the last Full and all the Recovery Point jobs and restore all the way up to the last completed transaction log which was closed before the time selected. If that last transaction log was closed by Exchange at say 9:36am, you would have everything restore up to 9:36a even though you said 9:42am. This is for your protection since whatever caused a failure(ie: a Virus) may be contained in the next transaction log which crosses over the point of failure time. You can always restore individual items as needed. The more often a Recovery Point job runs, the more restore points you have to use.
To see that your CPS job is running correctly, and to understand why you cannot mix a CPS Exchange job and a regular Exchange job, look at your B2D folder in Windows Explorer. You will see IMG folders and within those you will see VDB folders. When the Full part of the CPS job runs, it will create one IMG folder for each storage group in exchange that you have. The database and log files are copied here and are dynamically browsed from here during a restore. Every time the Full runs, a new set of IMG folders are created. When a Recovery Point job runs, it will create a VDB folder under each IMG folder. Since Recovery Point jobs run more often than fulls, you will see numerous VDB folders after awhile under each IMG folder. Now, once you have a bunch of VDB folders, look at the first one and note the transaction logs contained within it. Then look at the second VD folder, then the third, etc. You will see that all transaction logs contained in the first folder, are in the second one plus some new ones, and all those are in the third one plus new logs, etc. This is why you do not mix CPS Exchange and Regular Exchange backups since the transaction log sequence has to be in order for CPS to work correctly upon restore. The VDB folders link everything together in between full backups. The transactions logs which appear to be duplicates between folders are actually just hard links pointing to the original copy. This is expected and is how things are tied together.
One thing to note, the hardlinks (an OS feature) will cause your byte count on the hard drive to appear a lot larger than it is. This is becuase each hard link reports itself to the OS as a real file, even though it is just a link. This will result in a hard drive which say has a 500gb capacity sometimes reporting it is using 850gb. As the IMG folders expire and you run out of diskspace, the oldest IMG folders are recycledd and new ones are written in their place. Just set your diskspace reserve on the B2D folder to give it as much space as you can spare and let it use it. The disk space reserve is ow much space on the drive you are eleving for OTHER programs, not the B2D device. If you have a 500GB drive and want to give 400gb to the B2D device, set the reserve at 100GB. Keep in mind, this does not guarantee 400gb to the B2D folder, it just reserve at least 100gb for other folders. Other B2D folders on the drive, or other folders created later will eat into what you gave the B2D folder.
There is no Continuous Proteecction for SQL at this time. Set up normal Full/Differential/Incremental jobs to protect your SQL database.
Anyways, hope this clarifies what CPS Exchange jobs are and how to set them up in your environment.
Message Edited by D B on 08-08-200706:53 AM
