Details
-
New Feature
-
Status: Open (View Workflow)
-
Minor
-
Resolution: Unresolved
Description
Currently some information about all nodes is available at once via computer/api, but O
followup requests are needed to get the full node config. In some cases it makes sense to include information from the config in computer/api (JENKINS-25286). But unless the whole config is piggy backed, there will always be something missing for someone.
By including a Last-Modified or hash of the config as a field in computer/api, it would be easy to for api clients to locally cache the node configs and avoid the O lookups
Attachments
Activity
Chris Burroughs
added a comment - There is a followup request per node to get the config.xml for each node. A client first needs to know what nodes exist, before it can fetch the config.
depth and tree control the information returned by computer/api, but the node configuration (computer/$NAME/config.xml) is not part of that api.
Chris Burroughs
added a comment - There is a followup request per node to get the config.xml for each node. A client first needs to know what nodes exist, before it can fetch the config.
depth and tree control the information returned by computer/api, but the node configuration (computer/$NAME/config.xml) is not part of that api. That's true.
FWIW there's no reason this should be limited to nodes, if implemented. Anything with config.xml should be adapted accordingly.
It's not clear what this is asking for. Don't depth and tree solve the issue here? Why are there O(N) followup requests?