I have a few questions:
a) when you mention custom user type attributes this is regarding the User object and the IDC, right?
b) custom attributes that are ACM managed won't have negative impact, right?
c) custom attributes (collected, for any object) only have negative impact if the collector actually uses/fills them. E.g. a custom attribute for an entitlement might get used by EDC1 but EDC2 does not provide any data. Hence EDC2's performance won't be affected by the existence of that custom attribute?
PS: as a side note: it would be great if the circuit breakers for a multi-application collector would trigger for each "sub" collector and not for the change % of the overall data volume.
a) Yes a User Type Attribute that resolves to a user object. The overhead is that we collect the "raw" data and for each attribute we then need to identify which User it is. So if the customer has 5 attributes that requires 5 passes through the new/changed data. This is similar to the pre-70 collectors as well, but I wanted to call it out as the potential for impact.
b) That is correct a ACM Managed attribute is ignored by the collectors.
c) So any custom attribute, other than a User Type, is just data to the collectors that we store if we collect something. There is no additional overhead as it is just another column of data that we copy to the database. The User Type we still attempt to process even if we did not collect any raw data for the attribute. We don't really know if there was any User information collected until we try to process that attribute.
PS: We have gotten an RFE for that and it does make sense to have something like that.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.