-
Notifications
You must be signed in to change notification settings - Fork 93
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
Allow to independently monitor and dump events. #57
base: master
Are you sure you want to change the base?
Conversation
In large networks there is much traffic on the socket. This induces high load even when only a subset of the data is relevant. This commit introduces new commands on the socket for monitor neighbour, monitor route, monitor xroute, monitor interface that will exclusively show changes for these structures.
I like your changes. Thank you for bringing it up again. |
What is the addition supposed to achieve?
That's way to verbose. Please say something like
Please say instead
There are many occurrences of this style. |
If my understanding is correct, the "dump" (code 2), "monitor" (code |
FYI: I'm maintaining the opposite OpenWrt IPC ubus Part. ;) I plan something like this, too. |
Hi! |
@fcladera this allows to monitor/dump certain events like exclusively route changes reducing the CPU usage compared to monitoring every event significantly on busy networks. The added manpage bits should already explain how to use it. If it doesn't, please provide feedback what might be missing. @jech what is your opinion on the code? Can we get this merged? Using babeld in large networks on cheap plastic routers is CPU bound - not just because of babeld but also because of the processing load of surrounding daemons that parse events from babeld. It would be nice to be able to lighten the load a bit. |
any news on this one? I think @MisterDA addressed the concerns raised in the discussion. It would be nice to reduce CPU load by filtering messages before they are emitted. |
@jech Have you found time to check out the latest changes? It seems the code is ready to merge. |
Also, I think the code around line 1163 is deeply suspicious: the way we set I suggest reworking this code to parse both keywords (e.g. monitor route) in a single function. |
That looks fragile to me, you need to count. What about
and so on. |
This code was written by @christf some time ago. I forgot about it, but now it’s here so we can keep track and see what to do.