mononoke: be less specific in errors we expect in test-lfs-server-scuba-logging.t

Summary:
Like it says in the title. This isn't giving us the same error consistently
causing flaky failures.

Reviewed By: StanislavGlebik

Differential Revision: D28091747

fbshipit-source-id: dfc7a28b443c6577823c71cee7b006ed30fec18e
This commit is contained in:
Thomas Orozco 2021-04-29 06:43:33 -07:00 committed by Facebook GitHub Bot
parent f8000eeb86
commit 71994ae65d

View File

@ -258,12 +258,13 @@
Caused by:
invalid blake2 input: need exactly 64 hex digits
# Send a request after corrupting our data, and check that this gets logged too
# Send a request after corrupting our data, and check that this gets logged
# too. Silence the error we get so that output variations in Curl don't break
# the test.
$ truncate -s 0 "$SCUBA"
$ find "$TESTTMP/blobstore_lfs1" -type f -name "*chunk*" | xargs rm
$ curl -fsSL "${lfs_root}/lfs1/download/d28548bc21aabf04d143886d717d72375e3deecd0dafb3d110676b70a192cb5d" -o /dev/null
curl: (18) transfer closed with 2048 bytes remaining to read
[18]
$ curl -fsL "${lfs_root}/lfs1/download/d28548bc21aabf04d143886d717d72375e3deecd0dafb3d110676b70a192cb5d" -o /dev/null || false
[1]
$ wait_for_json_record_count "$SCUBA" 1
$ jq -r .normal.error_msg < "$SCUBA"
Chunk not found: ContentChunkId(Blake2(1504ec6ce051f99d41b82ec69ef7e9cde95054758b3e85ed73ef335f32f7263c))