Extend search timing observability to node, search, and ssh commands

[nikhil2611]

Epic

Part of Crawl Track backlog epic.

Background

Ex9 (Observability) added a Process::CLOCK_MONOTONIC timing hook only to knife status. The same pattern should be applied to other high-latency commands.

Code Paths

• lib/chef/knife/status.rb:66 — reference implementation (timing around q.search)
• lib/chef/knife/search.rb — #run method wraps Chef::Search::Query
• lib/chef/knife/node_list.rb — iterates all nodes via server API
• lib/chef/knife/ssh.rb — search + SSH fanout

Acceptance Criteria

• knife search logs: knife search: query returned N result(s) in Xs
• knife node list logs: knife node list: fetched N node(s) in Xs
• knife ssh logs: knife ssh: search returned N node(s) in Xs
• Log level is Chef::Log.info (same as status hook)
• New specs verify log output for each command

Dependencies

• Depends on: Ex9 observability hook (already merged)
• Blocks: nothing

Estimated Size

Small–Medium — one PR per command