Statd: improve speed of interface data #651
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This is a major redesign of statd interface handling. The goal here is to reduce the time it takes for statd to produce operational data for all interfaces on larger iron. Which manifested itself most notably in the CLI command "show interfaces". This patch removes the netlink interface subscriptions done by the statd c code. Instead statd now subscribes to the ietf-interfaces base path which means it will be invoked for all interfaces. If the user (sysrepo) has specified an explicit interface name such as eth0, it will be extracted by statd and passed down to Yanger. This patch also modifies Yanger to fit this new modus. No explicitly passed interface name now means all interfaces. Side note: The main culprit of the slow operational interface data is the time it takes to bootstrap python. Which prior to this patch was done for each interface. On the HW system I'm testing this patch on, runtime is reduced by a factor of 10. Signed-off-by: Richard Alpe <richard@bit42.se>
A non functional change. The caller knows more about how to format the args, so it makes more sense to build the Yanger cmdline from the caller. Signed-off-by: Richard Alpe <richard@bit42.se>
Debug print the full xpath in all callbacks. Signed-off-by: Richard Alpe <richard@bit42.se>
This is basically a functionally revert of: 435c5ca statd: allow ethtool --json to fail without breaking operational ds Having a fail flag that hides all yanger faults is too big of a hammer. If some hardware doesn't support a particular command such as "ethtool --json" it should be handled gracefully by yanger. Perhaps by catching the error and returning an empty dataset for that node. Signed-off-by: Richard Alpe <richard@bit42.se>
troglobit
approved these changes
Sep 25, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR removes the statd interface netlink code and improve speed of operational interface data.
This is a major redesign of statd interface handling. The goal here is to reduce the time it takes for statd to produce operational data for all interfaces on larger iron. Which manifested itself most notably in the CLI command "show interfaces".
This PR removes the netlink interface subscriptions done by the statd c code. Instead statd now subscribes to the ietf-interfaces base path which means it will be invoked for all interfaces. If the user (sysrepo) has specified an explicit interface name such as eth0, it will be extracted by statd and passed down to Yanger.
This PR also modifies Yanger to fit this new modus. No explicitly passed interface name now means all interfaces.
Side note
The main culprit of the slow operational interface data is the time it takes to bootstrap python. Which prior to this PR was done for each interface. On the HW system I'm testing this PR on, runtime is reduced by a factor of 10.