Membase crashes every few seconds
I've installed a fresh 1.6.5 two-node cluster for testing.
Everything was fine in the beginning. Then I wanted to test what happenes when I take down a node. The first few tries worked well (using the membase.server init.d script to start/stop).
However, after 2-3 restarts, one node started to issue "Service memcached exited on node 'ns_1@ipaddress' in 1.35s - supervisor_cushion001" every 6-7 seconds, and refused connections. Restart didn't help. I tried to restart the other node.... and from then on, even that other node showed the same symptoms.
In total, I stored only a few test items. One of the items I stored with an expire time, and occasionally I also got this error:
"Port server memcached on node 'ns_1@ipaddress' exited with status 134. Restarting. Messages: Item was expired at load: test
memcached: stored-value.hh:974: add_type_t HashTable::add(const Item&, bool, bool): Assertion `v->isDirty() == isDirty' failed."
So now I have a totally corrupted cluster, and I'm very glad it's only a test cluster. However, it seems to be a VERY bad idea to use a database which gets corrupted so easily in production. Is this some known bug?