Binlog checkpoint segfault

One of our servers inside a 3-node galera cluster is having troubles crashing itself around rsyslogd rotation (6:25am). We've also seen similar crashes sometimes when doing a full SST, after the data is transferred - but this is not consistent, we can restart the node fine most of the time.
It seems to only happen on this node, the other two nodes are working fine. All three nodes have the same hardware and software setup.

The full log has been attached.