fix(mojaloop/#3750): add timer for party lookup in cache

[kleyow]

fix(mojaloop/#3750): add timer for party lookup in cache - mojaloop/project#3750

• Change isLevelEnabled functions to object variable check for standardization between central logger and sdk logger feat(mojaloop/#3759)!: change logger functions to object variables sdk-standard-components#196
• Add mitigation for cache unsubscribtion edge case when performance characterizing transfers.

Details below

// The unsubscribeTimeout is used to mitigate a believed issue happening in the
// parties lookup leg of transfers. When the same party is looked up multiple times in quick
// succession, the cache is subscribed to the same channel multiple times. We believe that
// requests that have just subscribed to the channel which have not received the message yet
// are getting unsubscribed when requests that have completed call `unsubscribe`.
// This leads the request state machine to timeout the request, fail and
// stall the service. This issue is only affects parties lookup since it is the only
// pub/sub that can use the same channel name, primarily in our `ml-core-test-harness` environment.

This prevents an unsubscribe happening if a party is being constantly looked up until the requests cools down which inturn stops from the statemachines hanging then returning a 504.
Repeated party lookups for a single payee are unlikely to happen in a real world scenario but is prevelant in performance testing.

I researched methods of maybe trying to lock channel pub/sub, disregarding a rewrite of the cache implementation this is the easiest solution for the specific edgecase.

[sonarcloud]

Quality Gate failed

Failed conditions
16.1% Duplication on New Code (required ≤ 3%)

See analysis details on SonarCloud