feat(new source): Initial logplex source implementation
#1540
Conversation
Signed-off-by: Luke Steensen <luke.steensen@gmail.com>
binarylogic
changed the title
logplex source implementation
src/sources/logplex.rs
Outdated
|
|
||
| if events.len() != msg_count { | ||
| error!(message = "Parsed event count does not match message count header", event_count = events.len(), %msg_count); | ||
| debug_assert!(false, "Parsed event count does not match message count header"); |
There was a problem hiding this comment.
Can you explain why we need this?
There was a problem hiding this comment.
If we're in this branch, that means we've failed to parse the body correctly. In development, I want that to blow up, fail tests, etc. In production, I want to spit out an error but keep going so that we remain available for users.
There was a problem hiding this comment.
Would
if cfg_is!(test) {
panic!(...)
}be more clear here than debug_assert? Mostly, so that people know this should panic in tests.
| .map(|_| warp::reply()) | ||
| }); | ||
|
|
||
| let ping = warp::get2().and(warp::path("ping")).map(|| "pong"); |
There was a problem hiding this comment.
is this required or was this just for testing?
There was a problem hiding this comment.
Not required, but it's convenient when setting up infrastructure to make sure the server is reachable. Think of it as a minimum viable healthcheck, since we have no other routes without side effects.
src/sources/logplex.rs
Outdated
| } | ||
|
|
||
| #[test] | ||
| fn router_log() { |
There was a problem hiding this comment.
Would it make sense to test line_to_event a bit more? I wonder if we could fuzz or quickcheck at the most extreme too (not now but in the future).
There was a problem hiding this comment.
Yeah, it would. I've deliberately kept it pretty simple for now and provided fallbacks for when assumptions aren't met, but it would definitely be a good thing to fuzz.
In the meantime, I'll push a couple of other simple unit tests for it.
Signed-off-by: Luke Steensen <luke.steensen@gmail.com>
Signed-off-by: Luke Steensen <luke.steensen@gmail.com>
binarylogic
added
type: feature
Nothing too fancy, just accepts text bodies over HTTP where each line is roughly syslog-formatted. We do some basic parsing but fall back to forwarding the raw line if anything seems off.