fixed readmes, local ref in bench settings

README.md

@@ -21,7 +21,7 @@ Want to make your app faster and don't want to spend on extra infrastructure ? [
 **HA-store** is a generic wrapper for your data queries, it features: 
 
 - Smart TLRU cache for 'hot' information
-- Request coalescing, batching and retrying
+- Request coalescing and batching (solves the [Thundering Herd problem](https://en.wikipedia.org/wiki/Thundering_herd_problem))
 - Insightful stats and [events](#Monitoring-and-events)
 - Lightweight, configurable, battle-tested
 
@@ -56,13 +56,11 @@ Name | Required | Default | Description
 resolver | true | - | The method to wrap, and how to interpret the returned data. Uses the format `<function(ids, params)>`
 responseParser | false | (system) | The method that format the results from the resolver into an indexed collection. Accepts indexed collections or arrays of objects with an `id` property. Uses the format `<function(response, requestedIds, params)>`
 uniqueParams | false | `[]` | The list of parameters that, when passed, generate unique results. Ex: 'language', 'view', 'fields', 'country'. These will generate different combinations of cache keys.
-timeout | false | `null` | The maximum time allowed for the resolver to resolve.
+store | false | `null` | A custom store for the data, like [ha-store-redis](https://github.com/fed135/ha-redis-adapter).
 cache | false | <pre>{&#13;&#10;&nbsp;&nbsp;limit: 5000,&#13;&#10;&nbsp;&nbsp;ttl: 300000&#13;&#10;}</pre> | Caching options for the data - `limit` - the maximum number of records, and `ttl` - time to live for a record.
 batch | false | <pre>{&#13;&#10;&nbsp;&nbsp;tick: 50,&#13;&#10;&nbsp;&nbsp;max: 100&#13;&#10;}</pre> | Batching options for the requests
-retry | false | <pre>{&#13;&#10;&nbsp;&nbsp;base: 5,&#13;&#10;&nbsp;&nbsp;step: 3,&#13;&#10;&nbsp;&nbsp;limit: 5000,&#13;&#10;&nbsp;&nbsp;curve: <function(progress, start, end)>&#13;&#10;}</pre> | Retry options for the requests
 
 *All options are in (ms)
-*Scaling options are represented via and exponential curve with base and limit being the 2 edge values while steps is the number of events over that curve.
 
 ## Monitoring and events
 
@@ -71,11 +69,9 @@ HA-store emits events to track cache hits, miss and outbound requests.
 Event | Description
 --- | ---
 cacheHit | When the requested item is present in the microcache, or is already being fetched. Prevents another request from being created.
-cacheMiss | When the requested item is not present in the microcache and is not currently being fetched. A new request will be made.
 coalescedHit | When a record query successfully hooks to the promise of the same record in transit.
 query | When a batch of requests is about to be sent.
 queryFailed | Indicates that the batch has failed. Retry policy will dictate if it should be re-attempted.
-retryCancelled | Indicates that the batch has reached the allowed number of retries and is now abandoning.
 querySuccess | Indicates that the batch request was successful.
 
 You may also want to track the amount of `contexts` and `records` stored via the `size` method.

tests/profiling/README.md

@@ -1,6 +1,7 @@
 # Benchmarking HA-store
 
 1- You'll need to generate a sample with your desired distribution. Run this [generator](https://github.com/fed135/zipfian-generator).
+The bench expectations are based on these settings: `alpha: 0.5, size: 300000`. You can find a copy [here](https://gist.github.com/fed135/56282783a4c13d87a7f89c178b5d08d7).
 
 2- Update the [settings](https://github.com/fed135/ha-store/blob/next/tests/profiling/settings.js) for the test, making sure to update the path for the sample file. Then, adjust the assertion values for your bench test.
 

tests/profiling/settings.js

@@ -1,6 +1,5 @@
 const {getAssets} = require('./dao.js');
 //const redisStore = require('ha-store-redis');
-const redisStore = require('../../../ha-redis-adapter');
 
 module.exports = {
     test: {
@@ -15,11 +14,12 @@ module.exports = {
         retry: { base: 1, step: 2 },
     },
     assert: {
-        completed: [90000, 200000],
-        cacheHits: [20000, 70000],
+        completed: [300000, 300000],
+        coalescedHit: [8000, 15000],
+        cacheHits: [35000, 45000],
         timeouts: [0, 0],
-        batches: [500, 4000],
-        rss: [90000, 120000],
-        avgBatchSize: [35, 50],
+        batches: [4800, 5300],
+        rss: [50000, 80000],
+        avgBatchSize: [45, 50],
     },
 }

tests/profiling/worker.js

@@ -33,7 +33,7 @@ function handleRequest(id, language) {
     let finished = false;
     const before = Date.now();
     setTimeout(() => {
-      if (finished === false) suite.timeouts++;
+      //if (finished === false) suite.timeouts++;
     }, 500);
     store.get(id, { language }, crypto.randomBytes(8).toString('hex'))
     .then((result) => {