Conversation
|
@jcvenegas PTAL All - going forward we shouldn't need a PR via cherry-picks, this is just to start the bootstrap of the stable branches. Not sure if our CI will have issues with pre-existing Fixes? |
|
@nitkon @Pennyzct @lifupan @sboeuf @jcvenegas @jodh-intel @bergwolf @grahamwhaley @amshinde PTAL and ACK, as I am taking your prior commits/PRs to create this stable branch. |
|
Build failed (third-party-check pipeline) integration testing with
|
|
PSS Measurement: Memory inside container: |
|
Nice job @egernst - this looks like a real brainache, so yes, glad we only have to do this once. Given that, I think it might be prudent for folks to scan the From that summary line that feels like a bugfix. Similarly, those look on the surface like bug fixes. Eric, do you still have details of those in your head? If not, we can go read up on them to check.
/me wonders if we ever have to do this in the future (and I hope not ;-) ), if we can format the list with github tickbox stylies... with a little bit of sedding I reckon we could get to something like:
which might be easier to eyball-parse. Oh, and an added bonus, it appears to hot-link the SHA's! |
|
+1 on @grahamwhaley's comments. @egernst - since you've been through the list and added the "NO"'s manually, could you add a brief rationale for each since changes like... ... superficially look like they could/should be included. |
|
Now we have hotlinks, it is easy to check them...
Looks like a bugfix to me. We can debate how common it is maybe.
Also feels like a bugfix. Also up for debate. |
egernst
force-pushed
the
stable-1.1.1-candidate
branch
from
d91848f to
ce05b29
Compare
|
PSS Measurement: Memory inside container: |
|
PSS Measurement: Memory inside container: |
Codecov Report
@@ Coverage Diff @@
## stable-1.1 #621 +/- ##
==============================================
+ Coverage 64.24% 65.12% +0.88%
==============================================
Files 87 80 -7
Lines 8837 8242 -595
==============================================
- Hits 5677 5368 -309
+ Misses 2554 2325 -229
+ Partials 606 549 -57 |
|
Build failed (third-party-check pipeline) integration testing with
|
|
re: formatting: good idea -- I wasn't thinking of mark down - I was editing a snipped of git log --oneline --decorate --graph as I stepped through. I'll update the original paste for my rationale... |
|
PSS Measurement: Memory inside container: |
|
I added some notes above. @grahamwhaley I ended up pulling in the roll back. I was hesitant at first as there were a couple of conflicts introduced, but these were easily resolvable. Agreed. @jodh-intel, regarding @bergwolf - can you please weigh in? |
|
Build failed (third-party-check pipeline) integration testing with
|
|
PSS Measurement: Memory inside container: |
|
Build failed (third-party-check pipeline) integration testing with
|
|
@egernst According to the guideline we have agreed upon
I don't think we should include tests/logging related commits in stable. Code cleanups and New features should be excluded as well. If we follow this rule, it will cut down your current list quite a lot. |
|
Why not "tests/logging" ? I think it doesn't hurt anything. @bergwolf |
|
@WeiZhang555 I was echoing the guideline in the gdoc. I agree it's good to have tests and logging there. But it needs to be mentioned in the gdoc as well so that everyone is on the same page. The current statement there says that we only want minimal changes to the stable. Also by logging, I would restrict it as adding a few lines of new logs. For larger changes like the opentracing support @jodh-intel is currently working on, and any logging changes that require cross-component upgrades in order to work, should be excluded IMO. Looking closer, this one Similar is Also there are several code cleanups listed in the PR and we should decide if they are suitable for stable as well. |
|
Speaking of the guidelines, @egernst would you please add the branching and stable maintenance guidelines in your gdoc to the kata documentation repo? Since we are starting to practicing it, it's better to make it a formal document :) |
If some errors occur after qemu process start, then we need to rollback to kill qemu process Fixes: #297 Signed-off-by: flyflypeng <jiangpengfei9@huawei.com>
If kata-agent doesn't start in VM, we need to do some rollback operations to release related resources. add grpc check() to check kata-agent is running or not Fixes: #297 Signed-off-by: flyflypeng <jiangpengfei9@huawei.com>
If error occurs after sandbox network created successfully, we need to rollback to remove the created sandbox network Fixes: #297 Signed-off-by: flyflypeng <jiangpengfei9@huawei.com> Signed-off-by: Eric Ernst <eric.ernst@intel.com>
When createNetwork failed, cleanup the NetNs if it created. Fixes: #508 Signed-off-by: fupan <lifupan@gmail.com>
egernst
force-pushed
the
stable-1.1.1-candidate
branch
from
a1a4827 to
b2b817e
Compare
|
Repushed with problematic commits removed - thanks @bergwolf. Since we now have a stable release (which is just 1.1.0), I was able to verify with the initial stable packages as well. I'll work on clarifying the backport candidates as we move from gdoc to .md. I think that we should allow for low-risk changes which enhance debug / usability, especially at this point. Definitely not higher risk / larger changes though, which creep towards really being features (like opentracing). WDYT? We can continue the discussion on the document PR... |
|
Build failed (third-party-check pipeline) integration testing with
|
|
@egernst Thanks for the update. |
|
Thanks for adding the explanations @egernst! lgtm |
egernst
force-pushed
the
stable-1.1.1-candidate
branch
from
b2b817e to
057361b
Compare
|
Build failed (third-party-check pipeline) integration testing with
|
|
I still see some license headers issues: |
|
@chavafg the file header is correct, I think is an issue with the CI script. |
|
kata-containers/tests#652 should fix the issue. |
|
Build failed (third-party-check pipeline) integration testing with
|
jcvenegas
force-pushed
the
stable-1.1.1-candidate
branch
from
1403e9e to
65af3ea
Compare
|
Build failed (third-party-check pipeline) integration testing with
|
egernst
force-pushed
the
stable-1.1.1-candidate
branch
from
65af3ea to
7f426fd
Compare
|
Build failed (third-party-check pipeline) integration testing with
|
Add the runtime CLI command name to the virtcontainers logger so that it is clear when reading virtcontainers log entries which runtime command they refer to. Fixes #448. Signed-off-by: James O. D. Hunt <james.o.hunt@intel.com> Signed-off-by: Eric Ernst <eric.ernst@intel.com>
egernst
force-pushed
the
stable-1.1.1-candidate
branch
from
7f426fd to
6374e40
Compare
|
Build failed (third-party-check pipeline) integration testing with
|
|
CI was not stable with CRI-O tests on 1.1 lets merge this PR. |
Some test fixes were introduced into the 1.9 and 1.10 branches of cri-o. These fixes will help us minimize random failures. Fixes #481. Signed-off-by: Salvador Fuentes <salvador.fuentes@intel.com>
|
Build failed (third-party-check pipeline) integration testing with
|
revert: agent: sandbox_pause should get arguments from proc
14 participants
Walked through patches between 1.1.0 to HEAD, 8bbd9b5 , searching for changes which are candidates for a 1.1.x stable release.
All changes applied were non-conflicting cherry-picks.
Bugs from new features as well as new features or vendoring changes were rejected for stable.
Summary of changes/decisions made given below (for each commit between 1.1.0 and HEAD):
NO - * | e35caa6 Merge pull request #553 from chavafg/topic/update-openshift-version
^ don't want to update versions after making a stable unless there's an explicit fix we need
NO * | dec385a Merge pull request #548 from bergwolf/cleanup
^ Feature not relevant for 1.1.1
NO * | 244917c Merge pull request #546 from bergwolf/status-factory
Feature not relevant for 1.1.1
^VM factory changes not part of 1.1.1
NO * | 16600ef Merge pull request #531 from WeiZhang555/bugfix
^ refactoring which introduces risk - not a bug fix/security issue
NO * | | 04fa125 Merge pull request #544 from lifupan/master
^ factory not part of 1.1
NO * | | | c871a90 Merge pull request #539 from bergwolf/crng
^ Factory not part of 1.1
1.1.1 | * | | a8f5e2b (jodh/remove-initcall-debug-kernel-option) kernel: Remove initcall_debug boot option
1.1.1 | * | | | 476f21f codecov: remove codecov.yml
NO * | | | | 487f9ef Merge pull request #536 from bergwolf/qmp_clear
^ Not a bug fix/security issue
NO * | | | 1dae056 Merge pull request #524 from bergwolf/agent_log
^ Factory not a part of 1.1
NO * | | fc45d2e Merge pull request #497 from jcvenegas/vsock-runtime5
^New feature, not part of 1.1
NO * | | 17a2fb8 Merge pull request #528 from jodh-intel/fix-log-fields
^While "nice to have", this is changing an API, and modifies network code which was refactored since 1.1
NO * | f4a7712 Merge pull request #530 from kata-containers/revert-301-pre-addstorage-based-devmanager
^preadd storage Not part of 1.1
NO * | cfbc974 Merge pull request #521 from bergwolf/log
^factory not part of 1.1
NO * | | 2c3215c Merge pull request #301 from WeiZhang555/pre-addstorage-based-devmanager
^refactoring which isn't a bug fix nor security issue
NO * | | | 2a670ce Merge pull request #522 from chavafg/topic/update-docker-version
^ don't want to update versions after making a stable unless there's an explicit fix we need
1.1.1 | * | | | 5fc7219 agent: check: Increase timeout check request.
1.1.1 | * | | | 12e1911 kata-agent: Improve error message.
1.1.1 | * | | 4326ea8 cli: add AMD support to kata-check
1.1.1 | * | | 0e5f6b2 codecov: Explicitly disable codecov/patch coverage
1.1.1 | * | | c6fda44 virtconainers: rollback the NetNs when createNetwork failed
1.1.1 | * | | f9d5072 sandbox: change container slice to a map
1.1.1 | * | | 2993cb3 virtcontainers: fix kata-agent fail to start
1.1.1 | * | | 7103c4f virtcontainers: add qemu process rollback
1.1.1 | * | | c2651a8 virtcontainers: add kata-proxy rollback
1.1.1 | * | | daebbd1 virtcontainers: add rollback to remove sandbox network
NO * | | 99954d5 Merge pull request #501 from bergwolf/qemu
^ wasn't completely clear, but think this is just enabling for vm factory, not a bug fix/security issue
NO * | b244410 Merge pull request #505 from bergwolf/create_factory
1.1.1 | * 50b445c cli: tests: Clarify who cleans up tmpdir
1.1.1 | * 73c8286 cli: tests: remove the tmpdir to the config.json
1.1.1 | * d6d38da cli: update_test: defer remove tmpfile
NO * 14d25b8 Merge pull request #303 from bergwolf/vmfactory
^factory not part of 1.1
1.1.1 | * 5d6da35 kata_agent: print request details
NO * | 81c073f Merge pull request #458 from harche/without_devices
^new feature - not part of 1.1
1.1.1 | * c72a720 makefile: update PREFIX to remove redundant slash
1.1.1 | * bd6db30 kata-env: tests: add JSON out/in verify test
1.1.1 | * e45f591 kata-env: Do not leave temp files on test
1.1.1 | * 1f8f762 kata-env: Fix test cases for kata-env JSON
1.1.1 | * 63c06be kata-env: Add ability to output as JSON
NO * | 02056b6 Merge pull request #482 from chavafg/topic/update-cri-o
^ don't want to update versions after making a stable unless there's an explicit fix we need
1.1.1 031632d ci: no-exit: Do not run no-exit check on test files
1.1.1 62495d4 ci: no-exit: Allow path override for os.Exit check
1.1.1 6268ba4 ci: no-exit: Skip check if no files to check
1.1.1 1c453a3 ci: no-exit: Document and echo what the check is
1.1.1 | * | 114482e api: To stop its monitor after a sandbox paused
1.1.1 | * | d0bccab virtcontainers: Set ppc64le maxmem depending on qemu version
1.1.1 | * | a38b251 (jodh/pass-sandbox-to-proxy) tests: Refactor CC proxy test for Kata
NO:
^ changing compat
NO: * | a8952fb Merge pull request #470 from bergwolf/memory-hotplug
^ New feature
1.1.1 | * | 67de5d4 ci: reconstructure image-type and initrd for architecture-independant
1.1.1 830d50e (jodh/add-runtime-cli-cmd-to-vc-log) main: Pass runtime CLI command to vc logger
NO:
^ changing compat
1.1.1 64e2ca3 (jodh/improve-cid+sid-logging) logging: Reset external loggers when cid+sid available
1.1.1 03d4d49 logging: Add function to handle external loggers
1.1.1 79e8da0 create: Remove redundant logging code
1.1.1 9155412 api: To watch the vm console in FetchSandbox api