-
Notifications
You must be signed in to change notification settings - Fork 73
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
Bug: occasionally bogus response for mempool.get_fee_histogram() #197
Comments
the simplest reproduce via cli:
|
You're probably right there may be a bug. I'm looking at the code in Fulcrum and comparing it to ElectrumX: https://github.com/spesmilo/electrumx/blob/bbd985a95db63cada13254bb766f174e9ab674d0/electrumx/server/mempool.py#L154. I tried to just emulate their algorithm but I see they have since updated it. I will work on porting over the new changes to the algorithm. |
This should address issue #197. We synched the algo to latest fixes from spesmilo/electrumx.
Ok, I pushed this commit: 1ad892d -- this should address the issue. Note that I am traveling and thus may not be able to do a new release anytime soon to fix this. :/ Is there any way you can build Fulcrum yourself against master and test this and provide feedback that the issue is indeed solve? I was unable to reproduce because I don't have my server online right now. |
Ok well, 3 things:
|
.. it's also possible this is not a Fulcrum issue at all but rather the server itself is griefing its users for some reason or other (there are incentives to do this, of course, if one is also a miner). |
Yeah it's weird that the same server can return normal and faulty fees in consecutive requests. Probably need to reach out to Foundationdevices to investigate |
Hey, I run the fulcrum node at Foundation. This is indeed an odd situation. I want to reassure anybody coming across this issue that we are not involved in manipulating the fees reported by our nodes in any way. We don't mine or encourage our users to pay high fees unnecessarily. The reason we haven't seen this issue in the past is because we don't use fulcrum to report fees to our Envoy users, but rather mempool.space Some notes about our setup: we run two load-balanced fulcrum servers based on a "least connections" selection. We do this so we can safely update without downtime for our users by shifting traffic to one of the nodes, upgrading the other, and repeating that process until both are updated. So, maybe the weirdness seen where consecutive requests will return different results is due to that fact? Another thing is that we run our config with Anyway, apologies for the issues on our end. I will upgrade to |
@nickmonad Thanks for chiming in with more info on the issue. Sorry for insinuating it might be intentional on your part -- it's just such a mystery is all! I believe you that you aren't involved in manipulating fees and who knows maybe it is some esoteric bug. I do hope that the new code and the new setup will mean everything is resolved. Thanks for responding and being informative and helpful! |
@cculianu All good, thanks for cutting a new release with a possible fix. @Overtorment I ran your |
Well since I haven't heard any complaints about this continuing to happen.. will close this issue. If it happens can re-open. |
i was debugging unsusually high fees, and discovered that Fulcrum electrum server run by Foundationdevices occasionally returns bogus fee histogram, which leads to unusually high fees for the clients:
this is reproduceable occasionally (maybe its this specific server issue..?), and adequate response looks like this (that was run few minutes apart):
The text was updated successfully, but these errors were encountered: