Conversation
573e777 to
a59cf14
Compare
tests/util/util.go
Outdated
| } | ||
|
|
||
| func AssertBlobsEquals(a, b types.Blobs) { | ||
| // redundant for nice for debugging |
There was a problem hiding this comment.
I know this comment is from a copy/paste of previous code but is this check really redundant? Seems without the length check you would return true instead of false if there were extra blobs at the end of "b" that were not in "a".
There was a problem hiding this comment.
that's right. I'll remove the confusing comment.
| var Env *TestEnvironment | ||
|
|
||
| func InitE2ETest() { | ||
| if err := RestartDevnet(); err != nil { |
There was a problem hiding this comment.
Do we still need the RestartDevnet() function at all now that you remove this call? Seems to be the only call site.
There was a problem hiding this comment.
and then you could also get rid of StartDevnet() too I think.
| } | ||
| if err := ctx.Err(); err != nil { | ||
| return err | ||
| } |
There was a problem hiding this comment.
is it worth sleeping for say 1 second here?
There was a problem hiding this comment.
that might have been the issue in CI! I have another workflow running the 1 second patch that's succeeding!
Asserts that the beacon node follower can sync blobs during initial-sync. This patch also refactors the test harness to accommodate dynamic nodes.
Co-authored-by: Roberto Bayardo <bayardo@alum.mit.edu>
cafdabc to
636d8ff
Compare
* Add historical sync test (#44) Asserts that the beacon node follower can sync blobs during initial-sync. This patch also refactors the test harness to accommodate dynamic nodes. * update subrepos to latest Co-authored-by: Inphi <mlaw2501@gmail.com>
…ongoing development (#60) * create devnet v3 branch including submodules + new submodule for erigon client * Add Lodestar * Supply chain config * Cache all docker work before yarn build * Build packages individually to allow them to cache * We must also build the CLI * Fix clean * Update to head of Lodestar * Fun through Capella * Latest Lodestar * latest prsym * Remove nonfunctional setting * Bump Lodestar * update geth to latest merge w/ master (#46) * Make all beacon chain config values explicit and move to common directory (#48) * Add common beacon chain config file * Move PRESET_BASE to top of config file * bump service start timeout from 1 to 2 min * Revert timeout change from 2 minutes to 60 sec Co-authored-by: Roberto Bayardo <bayardo@alum.mit.edu> * add BLST_PORTABLE to CGO_CFLAGS to fix SIGILL issue with blst (#49) * Update lodestar submodule. Copy trusted setup * Update devnet v3 with latest subrepos and new test from master (#51) * Add historical sync test (#44) Asserts that the beacon node follower can sync blobs during initial-sync. This patch also refactors the test harness to accommodate dynamic nodes. * update subrepos to latest Co-authored-by: Inphi <mlaw2501@gmail.com> * Update test harness to start either prysm or lodestar * Fix the Prysm communication * Fix the Lodestar wiring * Fix up validator <-> beacon node communication * Start switching tests to real beacon API * fee-market tests pass against Prysm without gRPC * Fix initial-sync test * Switch prysm to my branch which has GetBlockJSON * Configure Prysm to upgrade later (as though it had Capella). Also run Lodestar pre-4844 test in CI * Fix make lodestar-up * Remove Prysm-specific chain config. We used a shared config now. * Point Prysm back at the devnet-v3 branch * Update geth to latest from devnet-v3 * Put erigon back * Remove preset base, let it take the default * Attempt to parallelize CI * idunno lol * Try two jobs * Comment failing Lodestar CI test * Fix clean * Use basename pwd instead * Go back to using --project-name, but correctly supply it to docker compose down also * upgrade subrepos to latest (#56) * update geth & prysm subrepos to latest, pointing at eip-4844 instead of devnet-v3 (#57) * update subrepos to latest, move back to using eip-4844 branch for development * add 30 min timeout to some tests that sometimes hang * fix CI, Makefile, and update Prysm & geth subrepos Note: the new prysm & geth subrepos use a different kzg trusted setup matching, which now matches the latest consensus spec Co-authored-by: dancoffman <coffman@coinbase.com> Co-authored-by: Jimmy Chen <jchen.tc@gmail.com> Co-authored-by: Inphi <mlaw2501@gmail.com> Co-authored-by: Daniel Coffman <dgcoffman@gmail.com>
Asserts that the beacon node follower can sync blobs during initial-sync.
This patch also refactors the test harness to accommodate dynamic nodes.