Method behind the madness

[eswbitto]

Hey fcorti,

I'm going through each report that comes with the AAAR install and there are a couple of questions I have on the query results defined by the scripts that are running.

In particular:

audit_details.prpt
audit_creation.prpt
folder_types.prpt


The results come back with data that has 'unknown' or 'surf-config'. My question is what are these results suppose to tabulate? To me they don't represent anything as far as actual audits of documents. Which could make the reporting inaccurate. Thoughts?

Also, In the login report I've tested on a fresh install of Alfresco and had only logged in once with the main admin account and when I ran the report I got a result of about 83 logins from the admin account. My assumption is that it is counting server side processes that alfresco does with the database using the admin account. (Is that right?)

I wonder if there is a way to get an actual login count from that user. If not I would just have to exclude that account from audits altogether.

[fcorti]

Hi ESWBitto,

The 'unknown' and 'surf-config' are probably related with the accessed documents.
The 'unknown' value is a "dummy value" that is used when the action is not referred to any document: for example a login (failed or not).
In this way you can choose to include or exclude those kind of actions from your analytics. 😉

For the 'surf-config'… probably some kind of accesses to a wrong url?

About the 83 logins from the admin account: congratulations, you are one of the firsts that observe this behavior. 🙂
Those are the accesses during the extraction process (AAAR_Extract) because you (correctly) use the admin account.
My suggestion is to define a "analytic" user in alfresco (with administrator privileges) and use this in the AAAR extractions.
Using this user you could choose to ignore the accesses, developing a filter in the reports queries.

Good job ESWBitto: your questions indicates you are really using AAAR in the right way, focusing the work on the data quality.
I have some ideas about the check of the "data quality"… but I don't know when they will be ready. 🙂