[LUCENENET-610] Reduce locking in FieldCacheImpl::Cache::Get

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: Lucene.Net 4.8.0
• Component/s: Lucene.Net Core
• Labels: up-for-grabs
• Assignee: Unassigned
• Reporter: Thomas Mathew
• Created: 12/Jul/19 22:37
• Updated: 09/Dec/21 18:01
• Resolved: 09/Dec/21 18:01

Description

We noticed a lot of contention in FieldCacheImpl::Cache::Get (our queries use a lot of query time joins + sorting, so we hit the field cache a lot).

We use a SearcherManager with warm-up queries to populate the field cache so we would expect it to be initialized in most cases before we hit it for actual requests.

The implementation seems to lock even for the happy path (when everything's already initialized). This seems like a by-product of the choice of data structures (the underlying WeakDictionary, WeakHashMap etc are not threadsafe) and so the locking is required in case the dictionary gets resized.

Ideally we could be using thread-safe data structures and only lock when initializing the data.

Issue Links

• Is contained by: LUCENENET-630 Identify/Fix Bottlenecks
• relates to: LUCENENET-640 Sequential IndexWriter performance in concurrent environments.
• relates to: LUCENENET-636 Port ConditionalWeakTable from .NET Core 3.x to .NET Standard 2.0