Check for existence of job before polling. Closes #330.

[ravwojdyla]

In case of local runners, export job does not exists, check if job
exists before polling for result. This will prevent unnecessary
exceptions, polling in local mode.

[googlebot]

Thanks for your pull request. It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

📝 Please visit https://cla.developers.google.com/ to sign.

Once you've signed, please reply here (e.g. I signed it!) and we'll verify. Thanks.

---

• If you've already signed a CLA, it's possible we don't have your GitHub username or you're using a different email address. Check your existing CLA data and verify that your email is set on your git commits.
• If you signed the CLA as a corporation, please let us know the company's name.

[ravwojdyla]

@peihe any comment on this one?

[dhalperi]

Pei: please take a look

[peihe]

I suggests to define the new function as:
/**
Gets the Job by {@link JobReference}.

Returns null if the job is not found. */ Job getJob(JobReference jobRef) throws InterruptedException, IOException;

The reason is: 1. getJob doesn't need to wait for the job to finish. 2. getJob is more general than exists.
Use the default BackOff for IOException:
new AttemptBoundedExponentialBackOff(MAX_RPC_ATTEMPTS, INITIAL_RPC_BACKOFF_MILLIS)

And, In the CleanupOperation, you can replace the pollJob() with getJob().

[ravwojdyla]

@peihe applied your suggestions.

[ravwojdyla]

ping @peihe

[peihe]

sorry for the delay. This PR looks good to me.
And, I appended some additional fixups in #351

Thanks for submitting the fix!

[peihe]

This RP is merged in #351

[ravwojdyla]

@peihe for some reason I can't find my commits in the history? Do you know what happen?

[ravwojdyla]

Ping @peihe

[dhalperi]

Hi Rav,

It looks like I messed up. GitHub automatically squashes pull requests into a single commit when merging, unless you specifically opt out. When I merged #351, which Pei sent as 3 commits, they got squashed fully into 1 commit.

On the command line, rebase+squash like this will result in git assigning authorship to the author of this first commit -- that should have been you. This is what we normally do when merging an external contributor's submission with minor fixups. However, it looks like GitHub may instead attribute authorship to the GitHub user who opens the PR, which in this case was @peihe.

Sorry for the gaffe. This difference in functionality is something I will watch out for this in the future.

At this point, we cannot retroactively repair the history. However, if you like I will revert (at HEAD) the squash and try to manually merge #351 again to properly attribute credit to you.

[ravwojdyla]

@dhalperi I see. No need for manual fix. Thanks.