Skip to content

Conversation

@jnewbery
Copy link
Contributor

@jnewbery jnewbery commented Aug 7, 2017

In v0.15, we disconnect nodes that send us version messages with
unsupported service bits (1 << 5 and 1 << 7). This commit adds a test
that bitcoind will disconnect those nodes before August 1st 2018, and won't
disconnect those nodes after August 1st 2018.

@sdaftuar @TheBlueMatt

In v0.15, we disconnect nodes that send us version messages with
unsupported service bits (1 << 5 and 1 << 7). This commit adds a test
that bitcoind will disconnect those nodes before August 1st 2018, and won't
disconnect those nodes after August 1st 2018.
@laanwj laanwj added the Tests label Aug 7, 2017
@jnewbery
Copy link
Contributor Author

jnewbery commented Aug 7, 2017

@laanwj - should be tagged v0.15 since it's a regression test for #10982

@TheBlueMatt
Copy link
Contributor

TheBlueMatt commented Aug 7, 2017 via email

@maflcko maflcko added this to the 0.15.0 milestone Aug 7, 2017
@maflcko
Copy link
Member

maflcko commented Aug 7, 2017

utACK 5e35cd9

@maflcko maflcko merged commit 5e35cd9 into bitcoin:master Aug 7, 2017
maflcko pushed a commit that referenced this pull request Aug 7, 2017
5e35cd9 [tests] Test disconnecting unsupported service bits logic. (John Newbery)

Pull request description:

  In v0.15, we disconnect nodes that send us version messages with
  unsupported service bits (1 << 5 and 1 << 7). This commit adds a test
  that bitcoind will disconnect those nodes before August 1st 2018, and won't
  disconnect those nodes after August 1st 2018.

  @sdaftuar @TheBlueMatt

Tree-SHA512: cb1bf311f9edc4aa5acf621e392543fd92952b2462ae2a8c20339e5c41e67ece1a2a9ede38db5dcd16563faa9ee58318e118f54024dd282c580132202d2df01b
@ghost
Copy link

ghost commented Aug 9, 2017

Why not relax the rule to instead allow for a maximum of e.g. 3 connections to nodes with the service bits 5 and 7? That way the network may stay fully connected both before and after a 2x HF in three months.

@maflcko
Copy link
Member

maflcko commented Aug 9, 2017

This was meant as a minimal fix/diff to be included in the first release candidate of 0.15.0, which was already due last Sunday. Beside the additional review required for more complex changes to p2p logic, it is unlikely that a more involved fix makes it into 0.15.0. Code cleanup for 0.16.0 is welcome, though.

@jnewbery jnewbery deleted the unsupported_service_bits_test branch August 9, 2017 19:26
@TheBlueMatt
Copy link
Contributor

TheBlueMatt commented Aug 9, 2017 via email

@jnewbery
Copy link
Contributor Author

jnewbery commented Aug 9, 2017

Please can we lock this PR? This PR added a new functional test case and is not the place to discuss the merits/demerits of the node disconnection behaviour in 10982. Thanks!

maflcko pushed a commit that referenced this pull request Sep 13, 2017
0063d2c [tests] Make p2p-leaktests.py more robust (John Newbery)

Pull request description:

  There has been an example of p2p-leaktests.py failing on travis in the new service bits test (introduced in #11001 . It appeared to me that the previous p2p connections had not been fully disconnected before attempting to add new p2p connections.

  I've added a sleep and restarted the NetworkThread, but I don't know whether this will fix the problem, since I'm unable to reproduce the failure locally.

  @MarcoFalke - not sure what you want to do here? I don't think this change could make things any worse.

Tree-SHA512: f5427c26267185a903c9b75bb3925bf153b8afce70c8e493bf8f585f57d809d20643b4ee69081300b211d22e960242aecc3d719f4ddd230aa08fdc5484b55055
PastaPastaPasta pushed a commit to PastaPastaPasta/dash that referenced this pull request Jan 10, 2020
0063d2c [tests] Make p2p-leaktests.py more robust (John Newbery)

Pull request description:

  There has been an example of p2p-leaktests.py failing on travis in the new service bits test (introduced in bitcoin#11001 . It appeared to me that the previous p2p connections had not been fully disconnected before attempting to add new p2p connections.

  I've added a sleep and restarted the NetworkThread, but I don't know whether this will fix the problem, since I'm unable to reproduce the failure locally.

  @MarcoFalke - not sure what you want to do here? I don't think this change could make things any worse.

Tree-SHA512: f5427c26267185a903c9b75bb3925bf153b8afce70c8e493bf8f585f57d809d20643b4ee69081300b211d22e960242aecc3d719f4ddd230aa08fdc5484b55055
gades pushed a commit to cosanta/cosanta-core that referenced this pull request Jun 30, 2021
0063d2c [tests] Make p2p-leaktests.py more robust (John Newbery)

Pull request description:

  There has been an example of p2p-leaktests.py failing on travis in the new service bits test (introduced in bitcoin#11001 . It appeared to me that the previous p2p connections had not been fully disconnected before attempting to add new p2p connections.

  I've added a sleep and restarted the NetworkThread, but I don't know whether this will fix the problem, since I'm unable to reproduce the failure locally.

  @MarcoFalke - not sure what you want to do here? I don't think this change could make things any worse.

Tree-SHA512: f5427c26267185a903c9b75bb3925bf153b8afce70c8e493bf8f585f57d809d20643b4ee69081300b211d22e960242aecc3d719f4ddd230aa08fdc5484b55055
@bitcoin bitcoin locked as resolved and limited conversation to collaborators Sep 8, 2021
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants