Who can tell me if I am right:
If the CAS is down the AMD buffers received data: until /var/spool is full, or the CAS comes back online. If the CAS is down for a limited periode of time no data is lost.
If the AMD is down, data is lost.
If you remove all data from the /var/spool on the AMD, the CAS does not loose any data because it is in his database. A small amount of data can get lost because it is not send to the CAS yet,
Does this make sense?
In general you're right, and only the first sentence should be commented.
AMD stores data files in /var/spool/adlex/rtm folder for defined number of hours (by default 192 = 8 days, configured in /usr/adlex/config/datacleaner.config) or until there is only 20% or 100MB of free space on HD (it's configurable in /usr/adlex/config/datacleaner.config, using "minFreeSize.percent" and "minFreeSize.value" properties).
The fact that CAS gets back online does not change it. So if your CAS is down for > 8days you will start loosing data.
Just as addition to what Adam stated - AMD will never preserve data for more days than configured.
The min. free space and min. % of free space can only shorten that period (in case when those exemplary 8 days of data occupy to much space).