I’ve had a 3-node cluster online for a few days, and I tried to take the second node down to change the tmpdir in my.cnf (disk was getting full). When I start MySQL now, the node will not recieve an SST and fails with some frustrating error messages. /www.e-six-sigma.net

First of all you should create this thread under “Percona XtraDB Cluster” category.
Regarding SST failure can you please provide mysql error log from the failed node along with my.cnf to identify root cause of the problem.