How to create Pool with tens of thousand nodes fast using REST

Interesting question... Is this for an existing pool or a new one? Tens of thousands of nodes in a pool? What requires that?

Gut feel tells me a config merge would be quickest... But I'm not sure if you can do that from REST... Never tried TBH...

H

I suspect you'd run up against some internal resource limits before you can get that far, not to say provisioning fast. The bigip.conf files do not seem to accept expandable variables, a mechanism used by, e.g. BIND in its named configuration files. It would be a good feature to have, though.

Up to a couple of thousand nodes in a Pool may be more practical, they can belong to a new Pool.

Once the Pool constructed, the change of node's membership may be more often than usual, so we are interested in any solution which confirms a committed change made to a set of nodes (e.g. up to a couple of hundreds) in one shot.

I might start looking at multiple layers of BigIP's myself at those kind of volumes...

e.g. 2 (HA'ed) at the front end, load-balancing say 8 others. Each of which has 1/8th of the servers (You could put them in a cluster and use 8x traffic-groups so that if one fails, one of the others can take over the load)... It's possibly not elegant, but at the volumes of nodes you're talking, it might be preferable.

Using VE's would help too...

H