Replace download_as_string with download_as_bytes

[jonathanmetzman]

download_as_string is deprecated.

[MjMoshiri]

Hi @jonathanmetzman, there was only one case and it's now replaced. 🙏🏻

[jonathanmetzman]

Thanks for pointing this out!

[jonathanmetzman]

Actually looks like it's still here https://github.com/google/clusterfuzz/blob/c2de16f9c9ae209a56156f69431e0bc48e608969/src/clusterfuzz/_internal/google_cloud_utils/storage.py#L289

[MjMoshiri]

Hello again @jonathanmetzman, pull request above changes that line but it's not approved yet.

[cdxvy30]

Sorry I am new to ClusterFuzz and trying to introduce it to my team. Does this issue mean that if I deployed a ClusterFuzz production setup with version 2.6.0 tag recently, reading files from Google Cloud Storage will not work?

I just deployed the ClusterFuzz to App Engine and can access to the .appspot.com, also built and uploaded OpenSSL fuzzer target to ClusterFuzz and the bot running on the GCP did show the string "fuzz libFuzzer libfuzzer_asan_linux_openssl". However, the ClusterFuzz homepage (ie: the Testcases page) did not show the testcase titled “Heap-buffer-overflow READ{*}” as expected. Seems like it did not fuzz the target I uploaded.

[jonathanmetzman]

This issue does not mean that. I'd look at the logs in google cloud to see why it's failed to produce crashes.