-
Notifications
You must be signed in to change notification settings - Fork 5.1k
ci: fix non-root regression, don't clobber Bazel cache on each invoca… #1915
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
…tion. Signed-off-by: Harvey Tuch <htuch@google.com>
This needs #1914 and the SHA update before merging. |
ci/build_setup.sh
Outdated
|
||
# Replace the existing Bazel output cache with a copy of the image's prebuilt deps. | ||
if [[ -d /bazel-prebuilt-output ]]; then | ||
if [[ -d /bazel-prebuilt-output && ! -d "${TEST_TMPDIR}/_bazel_${USER}" ]]; then | ||
BAZEL_OUTPUT_BASE="$(bazel info output_base)" | ||
rm -rf "${BAZEL_OUTPUT_BASE}" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure exactly what's going wrong with the old code, but if you want to avoid clobbering a fresher cache then this rm -rf
probably needs to change also.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It was symlinking in some stuff from the build image that Bazel complained about not being able to timestamp (?). Seems cleaner just to copy it rather than symlink to stuff generated by root on a read-only image.
The rm -rf
is OK, but probably redundant, as we now have a check above to see if the cache exists; if it does, we don't bother trying to copy over it. We can probably actually just remove this line completely.
LGTM |
Signed-off-by: Harvey Tuch <htuch@google.com>
@@ -1,3 +1,3 @@ | |||
# If you edit the SHA here you must also edit the SHA in .circleci/config.yml. | |||
|
|||
ENVOY_BUILD_SHA=44d539cb572d04c81b62425373440c54934cf267 | |||
ENVOY_BUILD_SHA=114e24c6fd05fc026492e9d2ca5608694e5ea59d |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oops, sorry, missed this. You need to update config.yml also. (See comment above). Do you mind doing a follow up? Just realized this when I merged master into my Lua branch.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Whoops. We should have some CI check that validates consistency here, or have the consumers of envoy_build_sha.sh
source from config.yml
via sed/grep. Either works for me, let me know what you prefer, I'll put in a PR to do this.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah, agreed. I'm open to any solution. sed/grep from config.yml is probably fine, or just some small check script that runs and checks both locations?
Signed-off-by: Harvey Tuch <htuch@google.com>
Before: $ bazel-bin/src/envoy/envoy --version bazel-bin/src/envoy/envoy version: 0/1.8.0-dev//DEBUG After: $ bazel-bin/src/envoy/envoy --version bazel-bin/src/envoy/envoy version: f315a32fc7c6f727fc9645cc1ca27d4160c1d0e0/1.8.0-dev/Clean/DEBUG Fixes envoyproxy#1803. Signed-off-by: Piotr Sikora <piotrsikora@google.com>
This is now default in Bazel 4.1, and the flag is removed in Bazel 5.0. Signed-off-by: Brentley Jones <brentleyjones@lyft.com> Signed-off-by: JP Simard <jp@jpsim.com>
This is now default in Bazel 4.1, and the flag is removed in Bazel 5.0. Signed-off-by: Brentley Jones <brentleyjones@lyft.com> Signed-off-by: JP Simard <jp@jpsim.com>
…tion.
Signed-off-by: Harvey Tuch htuch@google.com