Mutation request submitted twice for long running task - disable auto retry.

**Describe the bug**
Mutation request is submitted twice when backend takes a while to return the response or internet connection is slower.
It works when backend returns result quickly.

**How to disable mutation retry policy? i.e. submit mutation only once or fail.**

**To Reproduce**
Steps to reproduce the behavior:
1. Create AppSyncClient
```
private static volatile AWSAppSyncClient client;
public synchronized static AWSAppSyncClient getInstance(Context context) {
if (client == null) {
AWSConfiguration awsConfig = new AWSConfiguration(context);
client = AWSAppSyncClient.builder()
.context(context.getApplicationContext())
.awsConfiguration(awsConfig)
.build();
}
return client;
}
```
2. Submit mutation request
```
client.mutate(mutation)
.enqueue(new ...
```
3. Make sure backend executes a long-running task > 10-15 seconds.
4. The same mutation request is submitted twice.

**Expected behavior**
Mutations must not be submitted twice. Disable retry policy for mutations.

**Environment(please complete the following information):**
- AppSync SDK Version: [3.1.0]

**Device Information (please complete the following information):**
- Android 10
- Samsung A20

Start by tracing AWSAppSyncClient.mutate(mutation).enqueue(...) in the Android SDK and examine how a long-running request is handled after the 10–15 second delay described. Reproduce the issue with AppSync SDK 3.1.0 and confirm that a mutation is submitted only once or fails without a second submission.

Written by the indexing model from the issue text.