Security extraction from various IBM Connection components by IBM Content Analytics with Enterprise Search seedlist crawler for IBM Connections 4.5
If IBM Connections' Wiki component is down, then does IBM Content Analytics with Enterprise Search check for other components to extract user information from IBM Connection? If not, then does it mean that Connection secured search won’t work if wiki is part of the ICA index and if it’s down?
Resolving the problem
There are two similar cases:
Case 1- The crawler is NOT configured to crawl a Wikis feature.
If there is no Wiki component being crawled by the crawler, one of the other features is used for the group extraction. And this dictates the priority of how the API chooses a Connections feature among multiple features configured in the crawler. It is random (ie one of the other components is selected at random) unless the crawler is configured to crawl a Wikis feature. If the crawler is configured to crawl the Wikis feature, it always accesses to the group extraction API on the Wikis feature.
If the crawl spaces which are configured to the same host name and port contain Wikis feature, it has priority. Because Wikis feature has extra groups in addition to groups defined on other features. Otherwise, the API picks one of them randomly.
Case 2 - The crawler IS configured to crawl a Wikis feature, but the Wikis feature is temporarily down.
If Wiki is configured in the crawler as one of the components and if it goes down then, there's no fallback. ICA uses the API on Wikis feature, because it's different from others. The API on the other features cannot be a substitution.
So it means that Connections secured search won’t work if wiki is part of the ICAwES index and if it’s down.