Skip to content
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

Subscribers limited to one master at a time #127

Open
stonier opened this issue Dec 13, 2012 · 3 comments
Open

Subscribers limited to one master at a time #127

stonier opened this issue Dec 13, 2012 · 3 comments
Labels
Milestone

Comments

@stonier
Copy link
Member

@stonier stonier commented Dec 13, 2012

Masters send notify updates to subscribers when publisher connections go up and down. This wouldn't normally be a problem, but these updates aren't just notifying when publisher xxx goes up and down.

Instead it sends the complete local publisher list which the subscriber then synchronises to. This leaves it in a great state on the local system, but will delete all other connections currently present on remote systems as well.

End result, only one master has control at a time (whoever sent the last notification).

@piyushk

This comment has been minimized.

Copy link
Collaborator

@piyushk piyushk commented Dec 13, 2012

If I understand this correctly, there is no way around this problem without relaying information outside the ROS framework?

I assume that the original master_sync also suffered from this.

@stonier

This comment has been minimized.

Copy link
Member Author

@stonier stonier commented Dec 13, 2012

I experimented a little this morning. The gateways can do a better job of maintaining proper connections when flipping and unflipping via use of the node's slave api and a bit of information sharing.

However the real problem is caused by masters' sending updates when publishers locally come up and down - we have no control over this right now. So yes, no way around it without some updates to the RoS framework. On the flip side (pardon the pun), I think those updates could be relatively minor.

And yes, looking back at master sync, I expect this is why they only flip remote publishers locally and local publishers remotely. Obvious in hindsight.

@stonier

This comment has been minimized.

Copy link
Member Author

@stonier stonier commented Aug 5, 2018

See also the analysis in #125

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
2 participants
You can’t perform that action at this time.