I believe that this, in my testing seems to help the e2e state-sync
tests complete more reliably, by fixing some potential, range-related
slice building, as well as the way the test app hashes snapshots.
Additionally, and I'm not sure if we want to do this, but I added this
hook to the reactor that re-sends the request for snapshots during the
retry. This helps in tests prevent systems from getting stuck, but I
think in reality, it might create more traffic, and operators would
just restart a state-syncing node to get a similar effect.
also
- replace `MaxReconnectAttempts`, `ReadWait`, `WriteWait` and `PingPeriod` options with `WSOptions` in `WSClient` (rpc/jsonrpc/client/ws_client.go).
- set default write wait to 10s for `WSClient`(rpc/jsonrpc/client/ws_client.go)
- unexpose `WSEvents`(rpc/client/http.go)
Closes#6162
E2E tests often fail because validators miss signing or proposing blocks. Often this is because e.g. there's a lot of disruption in the network or it takes a long time to start up all the nodes.
This changes the test criteria to only check for 3 signed/proposed blocks, rather than a fraction of the expected blocks. This should be enough to catch most issues, apart from performance problems causing nodes to miss signing/proposing, but we may want separate tests for those sorts of things.
Executed a local network using simapp and looked for logs that seemed superfluous. This isn't by any means an exhaustive grooming, but should drastically help legibility of logs.
ref: #5912
E2E tests often fail due to fast sync stalls causing the validator to miss signing blocks. This increases the tolerance for missed signatures to 2/3 to allow validators to spend more time starting up.
closes: #5770closes: #5769
also, include node ID in the output (#5769) and modify NodeKey to use
value semantics (it makes perfect sense for NodeKey to not be a
pointer).