Updating tickets (#2316, #641, #1936, #1939, #1941, #1942, #1945, #2010, #2049, #2052, #2058, #2120, #2121, #2126, #2140, #2142, #2143, #2149, #2151, #2176, #2274, #1792, #1940, #1943, #1112, #1938)

Updating tickets (#1939, #1941, #1942, #1945, #2017, #2052, #2058, #2126, #2140, #2142, #2176, #2274)

Is this not something we need before really calling the cluster 'hardened'?
Sure, we could run it by hand a few times, but ideally this would be part of the continuos test suite.
At least I want to run some tests by hand with 100, 1000 and perhaps even more nodes so we can say that we have tested it on this.
So either raise the priority of this one or create a new ticket.

Sure, don't put too much into my rearrangement of priorities. It was just too much noise at the top of the list.

/Patrik

We can't test with that many nodes until we have addressed some of the known scalability related tickets. You have said that large cluster is not top prio. For me, after releasing the experimental memership, fd, and routers in 2.1, next big thing is the actor partitioning.

/Patrik

That is fine. But then we should test with as many as possible, if only 2 (which I don't hope is the limit), so we know where we are.

That is fine. But then we should test with as many as possible, if only 2 (which I don't hope is the limit), so we know where we are.