The basic answer is that the 'AL' command loads all samples and 'A' only loads samples relevant to the target application.
In more detail, I'll use an example of a fast running Monitor that a client created.
The job which ran was a DB2 batch IVP. It started at 15:16:30 and ended at 15:16:30 based on the TYPE=37 sample started record and the TYPE=2 termination record, respectively. So it ran in almost no time at all.
CA Mainframe Application Tuner (MAT) was able to initialize a DB2 packet for DB2 8.1, then MAT requested 8 observations, and ended up with an 11 count for observations meaning 3 times the DIE rescheduled itself because the job was gone. The job ran in about 120 milli-seconds and there wasn't a TCB anywhere that MAT could do anything with.
This is why when the "A" was attempted the empty monitor dataset message was received. There were records, but none that the Analyze Normal could use to do it's analysis.
If you do an "A" Analyze Normal, and receive the empty monitor dataset message, make sure you also do the "AL" Analyze All command to make sure that the monitor dataset is truly empty.
If you receive the empty monitor dataset in both instances, but believe that you should have samples to analyze, please contact CA Support Online and we will verify that your Monitor is indeed empty. We will create a FTP directory, and request that you TERSE the Monitor dataset and SFTP BINARY FB 1024 6144.