We have done a couple of Google integrations, some on the NetIQ platform and a few on other platforms. One common problem for
all of these is performance. One change (one API call) can take everything from a second to 30s to finish.
With NetIQ (the Concensus driver) is frequent backend errors (500 Internal Server Error). The performance and the fact that one
event might retry a couple of times due to backend errors makes it almost impossible to handle initial loads, larger re-syncs
and larger changes.
As an example, on the 2nd of November I re-synced 4000 accounts to Google to correct group membership. In ctp I query for
current memberships to make sure I don't do any unnecessary changes. The driver is still working and there is 1600 re-sync
events in the cache waiting...

I've googled, searched the forum and tried to find similar problems. There are quite a lot about different backend errors but
most of them seems to be on/off and solved by just doing right. There are very few describing scenarios where the same API call
fails for the first 3 times but then succeeds. On thought is that this might depend on which data centre google is using for the

Do any one else have the same problems? What is your best practice for a google integration at a larger customer?

Best regards,