4 ms·
> Yet threads make backtraces and control flow much less useful, since they miss out important context from concurrent threads. I mean, when a program panics,
by derefr 7y ago
> Yet threads make backtraces and control flow much less useful, since they miss out important context from concurrent threads.
I mean, when a program panics, you get a stacktrace from (a consistent snapshot of) all the threads, so what's the problem?
- erik_seaberg 7y agoIf a highly concurrent service gets a query'o'death, it's hard to tell which threads were working on it and which were merely within the blast radius. Frameworks tend to roll their own notion of "request context" without it being strongly typed or pervasive across the language and libraries.
- derefr 7y agoNow I’m curious what backtraces from Postgres look like when parallel scans are enabled. IIRC, you’ve got one isolated fork(2)ed master for the connection, which then has threads to divide work. Not too bad to debug.
- gardnerbickford 7y agoA snapshot doesn't show the shared memory changes that may have caused the panic. There needs to be a synchronous log of all changes to shared memory to debug an issue. Similar to having a snapshot of network traffic vs a recording of network traffic.