Kyle Kingsbury

Possible split-brain & loss of committed write with loss of un-fsynced data

11

### What happened? With a five-node etcd 3.5.3 cluster running on Debian stable, with a data directory mounted on the lazyfs filesystem, killing nodes and losing their un-fsynced writes appears...

Stale read with loss of un-fsynced writes

6

### What happened? In [this Jepsen test](https://github.com/etcd-io/etcd/files/8950246/20220621T110426.000-0400.zip) involving the loss of un-fsynced writes (via lazyfs) in a five-node cluster of etcd 3.5.3 nodes, we observed a stale read (a linearizability...

panic: unexpected error during txn

17

### What happened? In a Jepsen test run of five etcd 3.5.3 nodes, with process pauses (sending processes SIGSTOP and SIGCONT), compaction (performed via the admin API), and defragmentation (via...

Very occasionally, for reasons I don't understand (maybe TCP close?), a session disconnects between two calls to `ssh/invoke`. Works for one command, and a few ms later, another command throws:...

Okay, excellent. So, let's assume there's no easy way to optimize the width changes. But we can very easily optimize everything else. Strip all the jQuery calls out of partialRender...

Possible aborted reads with process crashes

2

**Versions** - etcd: 3.5.3 - jetcd: 0.7.1 - java: openjdk version "17.0.3" 2022-04-19 (Debian) **Describe the bug** I've got a Jepsen test case for etcd which appears to show [something...