When to requests were accessing the cache during the same time, there
was a race condition because of the concurrent access to the map which
holds the cached entries.
This is now fixed by adding a lock for the cache, so that multiple
requests can not access the cache in parallel.
When to requests were accessing the cache during the same time, there was a race condition because of the concurrent access to the map which holds the cached entries.
This is now fixed by adding a lock for the cache, so that multiple requests can not access the cache in parallel.
Fixes #157