-
Notifications
You must be signed in to change notification settings - Fork 92
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
client/eth/nodeclient_harness_test: First try with new harness fails. #1327
Comments
I'll look into this then. |
Thanks! Much appreciated |
oh wtf
So... sometimes light node can't find txs unless you intentionally pull the block first it seems.
How do we make this work if we don't know the block yet. Is this intended I wonder? |
Went ahead and made an issue ethereum/go-ethereum#24101 I think this is contributing to some of the test failures. Not sure why weren't seeing before though. Will wait for their reply to at least to see if this is intended. May even be due to geth updates that this is happeneing now where it wasn't before. |
Hmm, I thought that's what this was all about: ethereum/go-ethereum#19069 (comment)
So, I'm surprised too that you're seeing |
Maybe it's this race we talked about: #1317 (comment) |
I'll try twice. This bug wouldn't have anything to do with our code if it affects delta and gamma, they are just normal light nodes afaik. |
Yeah, I don't mean our code has the bug, just that a Go test could provide an easy way to step into the geth code handling the request. |
Is it intended that this doesn't return the error? dcrdex/client/asset/eth/nodeclient.go Lines 271 to 274 in fbff9b6
|
It was modeled after the internal/ethapi handler when we ditched the ethclient business for most calls: The reason why they return That doesn't mean it's right for our |
After calling this out I cannot even make it happen again. |
Oh shoo, you were right:
|
Closed with
so... "use full node" lol Probably @buck54321 has dug around in the geth lightnode code the most, and @chappjc, if you know what exactly is going on with fetching txs. It is able to fetch the tx after you ask for it. Also happens randomly. Seems like a simple race to me... Reposting chapps linked comment #1317 (comment) Could be a problem with the harness, which is best case. Can try to see if it happens on mainnet at all I guess. |
Ugh. Well leads we have:
At the end of the day, what's the consequence of pulling a |
On geth 1.10.15-stable now and haven't seen this is a while. Maybe fixed? |
I did just see:
|
Still not seen in a while. Tempted to close. |
Seems fixed, so closing. I think the last testGetCodeAt failure was fixed with new synced logic in #1630 |
The first run of tests always seem to fail in a few spots. Probably just needs some more time to sync? Subsequent tests use the same node directory which is probably why this happens on the first run, when the internal node is created.
Some error messages:
This last one is another testInitiate failure, but is kind of different than the others and may be something else.
The text was updated successfully, but these errors were encountered: