MCPX MCP Gateway

@legacy Feature: Strategy Based Queue @flakey Scenario: Delayed Processing with Requests That Timeout Given API Provider is up And Lunar Proxy is up When policies.yaml file is updated And policies.yaml includes a strategy_based_queue remedy for GET httpbinmock /anything/foo requests with 2 requests per 1 seconds and TTL of 2 seconds and queue_size of 10 resulting in 429 status code And policies.yaml file is saved And apply_policies command is run # And next epoch-based 1 seconds window arrives And 10 requests are sent in parallel to httpbinmock /anything/foo through Lunar Proxy Then requests 1, 2 are performed within window 1 returning status 200 And requests 3, 4 are performed within window 2 returning status 200 And requests 5, 6 are performed within window 3 returning status 200 # Since the queue's TTL is defined to be 2 seconds, # Any request which takes more than 2 seconds will result in 429 And requests 7, 8 are performed within 2 to 3 seconds returning status 429 And requests 9, 10 are performed within 2 to 3 seconds returning status 429 @flakey #Scenario: Drop requests when the maximum queue size is reached # Given API Provider is up # And Lunar Proxy is up # # When policies.yaml file is updated # And policies.yaml includes a strategy_based_queue remedy for GET httpbinmock /anything/foo requests with 1 requests per 10 seconds and TTL of 1 seconds and queue_size of 2 resulting in 429 status code # And policies.yaml file is saved # And apply_policies command is run # And 10 requests are sent in parallel to httpbinmock /anything/foo through Lunar Proxy # Then 1 requests returning with status 200 and 9 with 429 @flakey # Scenario: Prioritized Delayed Processing # Given API Provider is up # And Lunar Proxy is up # When policies.yaml file is updated # And policies.yaml includes a strategy_based_queue remedy for GET httpbinmock /anything/bar requests with 2 requests per 1 seconds and TTL of 5 seconds and queue_size of 10 resulting in 429 status code with prioritization of production > staging by header X-Env # And policies.yaml file is saved # And apply_policies command is run # And next epoch-based 1 seconds window arrives # And 8 requests are sent in parallel to httpbinmock /anything/bar through Lunar Proxy, 4 with X-Env header production and the rest staging # We can only assert on the non-immediate window, as no prioritization # occurs on the current window. # Then requests 3, 4 have production X-Env header # In this scenario, we can only be certain that staging requests take the last # two slots. # And requests 7, 8 have staging X-Env header # And requests 1, 2 are performed within window 1 returning status 200 # And requests 3, 4 are performed within window 2 returning status 200 # And requests 5, 6 are performed within window 3 returning status 200 # And requests 7, 8 are performed within window 4 returning status 200