Track node's XML ID rather than cluster ID for attribute write-outs in pacemaker-attrd #3796
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
... even if false, for code consistency and simplicity
... in write_attribute()
... for cleaner separation, and to ensure it is set whenever possible.
We only need the node XML ID for writing values to the CIB, so if a value will never be written, skip looking for the XML ID. This does mean that cluster nodes won't be added to the peer cache for unwritten attributes, but that shouldn't matter for them.
... to pcmk__cluster_get_xml_id(). It's getting the CIB XML ID; there's no such thing as a cluster-layer UUID, and it can be used with Pacemaker Remote nodes as well as cluster nodes.
... rather than using node->xml_id directly, so it gets set whenever possible. Also, make comparisons case-sensitive.
This effectively reverts 7afc160
Previously, attribute_value_t had a nodeid member to track the cluster ID of the node that the value is for. However the only reason we need it is to be able to get the node's XML ID in the CIB, for writing out the value. Rename it to node_xml_id and track the XML ID directly. In practice, there is no real change, since the CIB XML ID of Corosync nodes is simply their cluster ID as a string. This allows us to keep the same XML attribute and value in peer messages for backward compatibility. If we ever support node XML IDs that are *not* the string equivalent of their cluster IDs, rolling upgrades will be possible only from versions with this commit and later.
... for readability and to reduce code duplication
clumens
reviewed
Jan 10, 2025
Previously, node XML IDs were kept in attribute_value_t. That meant that they were duplicated for every value for a node, the values might be known for some values and unknown or inconsistent for others, and newly learned XML IDs would have to be broadcast per value. Now, maintain a global node XML ID cache.
|
This looks good to me now, but it would be great to get some other opinions before merging. |
|
Looks good to me too so far. Thanks! |
clumens
approved these changes
Jan 13, 2025
|
Added one commit on top to address review |
nrwahl2
approved these changes
Jan 14, 2025
Add a doxygen block, rename function to update_node_state_after_fencing() and uuid argument to target_xml_id for readability, and improve log messages, comments, and formatting.
nrwahl2
approved these changes
Jan 14, 2025
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.
No description provided.