Update README.md

Author: balis

examples/RemoteJobs/README.md

@@ -11,10 +11,10 @@ The distributed execution architecture consists of:
     - **Application software** - software that actually performs workflow tasks
 
 In this example:
- - The workflow invokes function `submitRemoteJob` (Job invoker) from `functions.js` 100 times. This function simulates submission of jobs by starting 100 parallel processes of `node handler.js  <taskId> <redis_url>`.
- - `../../../hyperflow-job-executor/handler.js` represents a remote job executor which is passed two parameters: `taskId` and `redis_url`. 
- - `handler.js` gets a `jobMessage` from HyperFlow, and then sends back a notification that the job has completed; `taskId` is used to construct appropriate Redis keys for this communication.
- - On the HyperFlow engine side, the Job invoker can use two functions (provided by Hyperflow): `context.sendMsgToJob` to send a message to the job, and `context.jobResult` to wait for the notification. These functions return a [`Promise`](https://javascript.info/promise-basics), so the async/await syntax can be used as shown in the example.
+ - The workflow has two tasks (see `workflow.json`): one that executes `job.js`, the other which simply runs `ls -l`. Note that the commands to be executed are specified in `workflow.json`. 
+ - The engine invokes the function `submitRemoteJob` (Job invoker) from `functions.js`. This function simulates submission of jobs by starting the Hyperflow job executor and communicating with it via Redis to run jobs.
+ - `../../../hyperflow-job-executor/handler.js` represents a remote job executor which is passed two parameters: `taskId` and `redis_url`. The executor gets a `jobMessage` from HyperFlow, executes the command in a separate process, and then sends back a notification that the job has completed; `taskId` is used to construct appropriate Redis keys for this communication.
+ - On the HyperFlow engine side, the Job invoker can use two functions (provided by Hyperflow): `context.sendMsgToJob` to send a message to the job executor, and `context.jobResult` to wait for the notification from the executor. These functions return a [`Promise`](https://javascript.info/promise-basics), so the async/await syntax can be used as shown in the example.
  - The parameter to the `context.jobResult` function is a timeout in seconds (0 denotes infinity). One can use a retry library, such as [promise-retry](https://www.npmjs.com/package/promise-retry), to implement an exponential retry strategy.
  - The Job invoker also gets the Redis URL in `context.redis_url` which can be passed to the remote job executors.