--- Log opened Tue May 03 00:00:13 2022 01:09 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 01:20 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Remote host closed the connection] 01:20 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 04:16 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Remote host closed the connection] 04:17 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 04:41 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Remote host closed the connection] 04:42 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 05:55 < darosior> I'm taking a stab at profiling the test suite, we can't go on with those spurious CI timeouts. wait_for_logs is indeed the culprit for a large part, but i suspect the coordinator polling on revaultd as well to be *very* inefficient 06:46 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Ping timeout: 240 seconds] 07:24 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 07:36 < darosior> I could reliably reproduce the spurious coordinator error, i'm fixing it now. 07:39 < edouard[m]> coordinator side or revaultd ? 07:44 < darosior> The failure to store the signatures on the coordinator side 07:52 < edouard[m]> The code is confusing for me, why moving the db connection in a thread::spawn ? 07:53 < edouard[m]> but if it fix the bug, I agree to merge the PR, for the long term we should have a connection pool 08:01 < darosior> Hmm? 08:01 < darosior> Ah that's existing code 08:02 < darosior> See the doc for the rationale for doing so: https://docs.rs/tokio-postgres/latest/tokio_postgres/index.html 08:02 < edouard[m]> cool thanks 08:07 < darosior> Yeah it fixes all the spurious "failed to store the signatures" error AFAICT 08:10 < darosior> So it would be nice to have in ASAP 08:18 < edouard[m]> utACK for me 08:21 < darosior> Yay! 08:31 < darosior> #394 reduces the runtime of test_large_spend, our test with the longest runtime, by 5 on my machine 08:31 < darosior> Now let's see for the entire test suite 08:57 < edouard[m]> I am trying #394 08:59 < edouard[m]> or not the ci is failing 09:00 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Ping timeout: 240 seconds] 09:03 < darosior> Yeah i spotted a few tests that are racy, and some timeout on the sig exchange now (probably the bitcoind thread is starving the sigfetcher one) 09:11 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 09:16 < darosior> edouard[m]: so #394 uncovered an issue with test_gethistory https://github.com/revault/revaultd/issues/396 09:28 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Ping timeout: 240 seconds] 17:03 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 17:08 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Ping timeout: 240 seconds] 21:43 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has joined #revault 22:09 -!- evanlinjin [~evanlinji@gateway/tor-sasl/evanlinjin] has quit [Ping timeout: 240 seconds] --- Log closed Wed May 04 00:00:14 2022