using sendtoaddress is calculating wrong/high fees

[richiela]

Using a standard sendtoaddress call during our priming phase is generating high txfees.

Couple of examples:

https://blockchain.info/tx/8bac7dc71a0753fe97c92adb3d3f5d8589555b12dabddd3c12c85fc506bdc7aa
https://blockchain.info/tx/c18110e7782762be8be0eb3bf5def66b5b85c014fc6929a6efeb00937e08e9b2
https://blockchain.info/tx/4a406dca52f7d89ffae80756cf371a749d70a087f878457a1eb32a781bdfaa14

This happened after we upgraded. The previous version we were using didn't see this behavior.

[halfcab123]

Has this been resolved ?

[richiela]

No, the issue still persists. We've been able to mitigate by hardcoding a value for settxfee.

[sipa]

This looks like it's the same issue as bitcoin#10247 upstream, which is actively being worked on.

[richiela]

That definitely sounds like it. Is there a plan to rebase to latest with the bug fixes?