Refresh with fresh sequence token

[mstovicek]

• What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)

Bug fix

• What is the current behavior? (You can also link to an open issue here)

An exception about an invalid CloudWatch log stream sequence token appears while running a long script in Docker. I suspect another process creates a new sequence token and invalidates the one used by a Docker instance. This will probably happen when running a cluster of few web instances trying to write logs in a same time too.

Exception:

23:03:37 ERROR [console] Error thrown while running command "*****". 
Message: "Error executing "PutLogEvents" on "https://logs.eu-central-1.amazonaws.com"; 
AWS HTTP error: Client error: `POST https://logs.eu-central-1.amazonaws.com` resulted in a `400 Bad Request` response:
{
  "__type": "InvalidSequenceTokenException",
  "expectedSequenceToken": "49588220171152916907562770980898637167375952589803759 (truncated...) InvalidSequenceTokenException (client): The given sequenceToken is invalid. The next expected sequenceToken is: 49588220171152916907562770980898637167375952589803759074"
}
{
  "__type": "InvalidSequenceTokenException",
  "expectedSequenceToken": "49588220171152916907562770980898637167375952589803759074",
  "message": "The given sequenceToken is invalid. The next expected sequenceToken is: 49588220171152916907562770980898637167375952589803759074"
}
{
    "exception" => Aws\CloudWatchLogs\Exception\CloudWatchLogsException {},
    "command" => "***",
    "message" => "Error executing \"PutLogEvents\" on \"https://logs.eu-central-1.amazonaws.com\"; 
    AWS HTTP error: Client error: `POST https://logs.eu-central-1.amazonaws.com` resulted in a `400 Bad Request` response:
    {
        "__type": "InvalidSequenceTokenException",
        "expectedSequenceToken":
        49588220171152916907562770980898637167375952589803759 (truncated...)
        InvalidSequenceTokenException (client): The given sequenceToken is invalid. The next expected sequenceToken is: 49588220171152916907562770980898637167375952589803759074
    }
}

• What is the new behavior (if this is a feature change)?

Handler retries to send entries to CloudWatch once an exception \Aws\CloudWatchLogs\Exception\CloudWatchLogsException is thrown by a client. It gets a fresh sequence token from CloudWatch log stream before retry.
It does not check the exact type of an exception simply because I did not notice another exception.

• Does this PR introduce a breaking change? (What changes might users need to make in their application due to this PR?)

No breaking changes

• Other information:

Hi Max,
we still use your library in our main project and we don't notice any troubles when running in for user requests 👍
I hope you will find a time to review and test this PR so we can upgrage to a new version.

[coveralls]

Coverage decreased (-3.9%) to 93.684% when pulling 14319c6 on buzzbird:retry-with-fresh-sequence-token into 4034b71 on maxbanton:master.

[coveralls]

Coverage decreased (-3.9%) to 93.684% when pulling 905f883 on buzzbird:retry-with-fresh-sequence-token into 4034b71 on maxbanton:master.

[maxbanton]

Hi @mstovicek! Thanks for your PR, merged. Will publish new version soon.

[maxbanton]

New version 1.1.10 published! https://packagist.org/packages/maxbanton/cwh#v1.1.10

By the way, the library recently scored 400K of downloads! Congratulations to the whole community!

[mstovicek]

Thank you for a very quick response, @maxbanton <3