Maximum practical number of databases clustered

I have seen the answer danced around but not answered. We have passed 10,000 databases per cluster member. I’m thinking we have too much fun, but all I can find is IBM stating in faq technote 171686, which is staged as if it were an answer to my question, that there is a 30 MB limit in R6 for internal clustering data. What data does the limit apply to? Must we ask the professor after class?

Also, there is a practical usability limit for things like how long maintenance tasks take, lookups in directories of databases, etc…

Do you think we’re past it? Has anybody got more than 10,000 NSF files on multiple cluster members?

PS

The domino admin “cluster analysis” tool has stopped working for us way back around 8 or 9000 databases. Too many false positives, useless data and timeouts. So there’s a practical limit already.

Subject: Maximum practical number of databases clustered

I opened an IBM ticket, and the tech hunted around. She finally found from conversations or internal technotes that whatever the practical number may be, it’s much larger than 10,000 because the internal limit was multiplied in releases several times since R5. She said it could be in the 100,000 range now.

Nevertheless, that’s no more practical than a 64 GB database. I need to just take a poll, I guess.