Cli change for file based request source #563

wjuan-AFK · 2020-10-22T19:35:55Z

Commit to add RequestSourcePlugin option to the CLI. base PR: #560.
This is necessary to make use of the previous PRs adding the requestsourcepluginfactory via the CLI.

Signed-off-by: William Juan <[email protected]>

a global request generator. Signed-off-by: William Juan <[email protected]>

Signed-off-by: William Juan <[email protected]>

Adding comments. Signed-off-by: William Juan <[email protected]>

Signed-off-by: William Juan <[email protected]>

…TestForRequestSourceFactory Signed-off-by: William Juan <[email protected]>

Signed-off-by: William Juan <[email protected]>

request source. Signed-off-by: William Juan <[email protected]>

Signed-off-by: William Juan <[email protected]>

repokitteh-read-only · 2020-11-03T01:45:00Z

🔨 rebuilding ci/circleci: clang_tidy (failed build)

🐱

Caused by: a #563 (comment) was created by @wjuan-AFK.

see: more, trace.

Signed-off-by: William Juan <[email protected]>

wjuan-AFK · 2020-11-03T05:09:20Z

/retest

repokitteh-read-only · 2020-11-03T05:09:24Z

🔨 rebuilding ci/circleci: clang_tidy (failed build)

🐱

Caused by: a #563 (comment) was created by @wjuan-AFK.

see: more, trace.

oschaaf

LGTM, thanks!
Leaving one last suggestion with respect to a test that now uses a textual yaml diff comparison as the sole means to test for equivalence, I suspect that doubling down with a true equivalence tests helps catch edge cases. (We really need #433)

oschaaf · 2020-11-03T09:50:07Z

test/options_test.cc

+  std::string s1 = Envoy::MessageUtil::getYamlStringFromMessage(
+      *(options_from_proto.toCommandLineOptions()), true, true);
+  std::string s2 = Envoy::MessageUtil::getYamlStringFromMessage(*command, true, true);
+  EXPECT_EQ(s1, s2);


Consider also adding the following comparison, as I suspect the yaml/text based comparison may miss edge cases.

EXPECT_TRUE(util(*(options_from_proto.toCommandLineOptions()), *command));

Ok. Understood. That makes sense.

@dubious90 Wanted to let you know that I was adding this back in since this reasoning makes sense to me.

Signed-off-by: William Juan <[email protected]>

wjuan-AFK · 2020-11-03T21:29:17Z

/retest

repokitteh-read-only · 2020-11-03T21:29:20Z

🔨 rebuilding ci/circleci: clang_tidy (failed build)

🐱

Caused by: a #563 (comment) was created by @wjuan-AFK.

see: more, trace.

dubious90

LGTM modulo a few commenting comments. Will continue tracking the two conversations there, but otherwise reassigning to mumak, who will approve.

dubious90 · 2020-11-03T21:22:54Z

test/options_test.cc

@@ -249,7 +250,7 @@ TEST_F(OptionsImplTest, AlmostAll) {
  EXPECT_TRUE(util(cmd->stats_sinks(0), options->statsSinks()[0]));
  EXPECT_TRUE(util(cmd->stats_sinks(1), options->statsSinks()[1]));
  EXPECT_EQ(cmd->latency_response_header_name().value(), options->responseHeaderWithLatencyInput());
-
+  // TODO(#433)


Two nit asks on this set of TODOs:

Can we add a oneliner, explaining what the action to do here specifically is

Rather than repeating it on every test here, we could have in the comment something like "Here and below". Then we only need to include the TODO once per relevant file

sounds good to me. I feel like tagging the comparisons which need to be replaced is helpful, but I can remove the other ones if you insist.

dubious90 · 2020-11-03T21:39:17Z

test/options_test.cc

+  CommandLineOptionsPtr cmd = options->toCommandLineOptions();
+  EXPECT_TRUE(
+      util(cmd->request_source_plugin_config(), options->requestSourcePluginConfig().value()));
+  // Now we construct a new options from the proto we created above. This should result in an


Sorry -- my previous code comment wasn't particularly helpful/directive. I agree that comments here would be helpful, but I'm not finding this comment to clarify the code for me, so let me re-explain better:
In general, comments are most helpful when we explain the why, rather than the what. This comment is actually doing a little bit of both. One of the reasons why I think you're finding the need to explain "what" is that the comments aren't actually positionally with the code that they're explaining.

Rather than explaining a whole series of actions in one comment, can we split up the comments to be directly tied to their code sections?

dubious90 · 2020-11-03T21:40:22Z

test/options_test.cc

+  OptionsImpl options_from_proto(*command);
+  std::string s1 = Envoy::MessageUtil::getYamlStringFromMessage(
+      *(options_from_proto.toCommandLineOptions()), true, true);
+  std::string s2 = Envoy::MessageUtil::getYamlStringFromMessage(*command, true, true);


One thing that might make this easier to parse is if we use more descriptive names, rather than s1 and s2. Maybe expected_options_yaml and actual_options_yaml? Or whatever names you think would be accurate/helpful.

dubious90 · 2020-11-03T21:46:51Z

test/options_test.cc

+INSTANTIATE_TEST_SUITE_P(
+    HappyPathRequestSourceConfigJsonSuccessfullyTranslatesIntoOptions,
+    RequestSourcePluginTestFixture,
+    ::testing::ValuesIn(std::vector<std::string>{


I'm having a really hard time parsing this. There are two things that might help here:

C++ has a formatted string option that will make this much easier to read, rather than using " so much. You can do R"( string that contains " with no problems)". https://abseil.io/tips/64

(optional) I think this is just one string, but it's difficult to tell. Can we abstract the actual textprotos into variables?

I'm not sure that 1 works for me here. I might be missing something, but it seems like there are issues with the newlines. At least it seems like using R"( makes you keep the newlines if you space them into separate lines. I can make each line a separate raw string though, I'm not sure that this makes it more readable, but maybe you'll find this preferable?
2. I've kind of tried to do this. Hope that this approach looks ok.

Signed-off-by: William Juan <[email protected]>

wjuan-AFK · 2020-11-05T16:26:54Z

/retest

repokitteh-read-only · 2020-11-05T16:26:57Z

🔨 rebuilding ci/circleci: test (failed build)

🐱

Caused by: a #563 (comment) was created by @wjuan-AFK.

see: more, trace.

mum4k

Looks good, no actionable comments from my end.

mum4k · 2020-11-05T19:44:14Z

@dubious90 this is good to go.

wjuan-AFK added 30 commits July 15, 2020 20:04

Adding a multiple path test to confirm that expected paths are hit

99f45d1

Signed-off-by: William Juan <[email protected]>

fixing asan problems

075a4cf

Signed-off-by: William Juan <[email protected]>

Fix Format

ecbf407

Signed-off-by: William Juan <[email protected]>

Removeing unused comments

8fb11b6

Signed-off-by: William Juan <[email protected]>

Removing unnecessary invoke calls for cleanliness

032dbba

Signed-off-by: William Juan <[email protected]>

Merge branch 'master' into RequestSourceUnitTestBenchmarkMultiplePath

953e953

Signed-off-by: William Juan <[email protected]>

Refactoring test to take in request generator parameter instead of using

6185edb

a global request generator. Signed-off-by: William Juan <[email protected]>

Changing to pass by reference.

cded6ad

Signed-off-by: William Juan <[email protected]>

Merge branch 'master' into RequestSourceUnitTestBenchmarkMultiplePath

fbfe523

Signed-off-by: William Juan <[email protected]>

Merge branch 'master' into RequestSourceUnitTestBenchmarkMultiplePath

0d70179

Signed-off-by: William Juan <[email protected]>

Refactoring per comments in PR. Moving helpers, renaming functions.

cdb7efe

Adding comments. Signed-off-by: William Juan <[email protected]>

Adding comments and renaming testbasicfunctionality for clarity

f15bb47

Signed-off-by: William Juan <[email protected]>

Consistent unit test naming convention using should. Clarifying comments

59359a2

Signed-off-by: William Juan <[email protected]>

Fix format

f63df91

Signed-off-by: William Juan <[email protected]>

Adding a clarifying comment for helper function.

406bd6b

Signed-off-by: William Juan <[email protected]>

Fixing typo and other nits

f0b697b

Signed-off-by: William Juan <[email protected]>

Renaming testBenchmarkClient for clarity

8425e48

Signed-off-by: William Juan <[email protected]>

Replacing Flat hash map with flat hash set for simplicity.

fb4f372

Signed-off-by: William Juan <[email protected]>

Comment clarification.

c6eae6c

Signed-off-by: William Juan <[email protected]>

Small refactor for factory method for more dependency injection.

96ae38b

Signed-off-by: William Juan <[email protected]>

Fix format.

b17240f

Signed-off-by: William Juan <[email protected]>

Refactor to inject the requestsourceconstructor.

5c5a146

Signed-off-by: William Juan <[email protected]>

Moving helper functions.

29591ff

Signed-off-by: William Juan <[email protected]>

Merge branch 'RequestSourceUnitTestBenchmarkMultiplePath' into Adding…

7d928a9

…TestForRequestSourceFactory Signed-off-by: William Juan <[email protected]>

Fix format.

ff82d95

Signed-off-by: William Juan <[email protected]>

Adding test using the new mock requestsourceconstructor to get remote

98381a4

request source. Signed-off-by: William Juan <[email protected]>

Fix format.

83ac556

Signed-off-by: William Juan <[email protected]>

Adding missing request source constructor files.

38c5084

Signed-off-by: William Juan <[email protected]>

Changing helper functions to anonymous namespace.

fa4e285

Signed-off-by: William Juan <[email protected]>

Changing to use a nullable pointer instead of reference for clarity.

8551c76

Signed-off-by: William Juan <[email protected]>

Comment clean up.

314f924

Signed-off-by: William Juan <[email protected]>

wjuan-AFK added waiting-for-review A PR waiting for a review. and removed waiting-for-changes A PR waiting for comments to be resolved and changes to be applied. labels Nov 3, 2020

wjuan-AFK requested review from dubious90 and oschaaf November 3, 2020 05:22

oschaaf previously approved these changes Nov 3, 2020

View reviewed changes

wjuan-AFK added 2 commits November 3, 2020 17:05

Adding back additional comparison to avoid edge cases missed.

587cf3c

Signed-off-by: William Juan <[email protected]>

Fix format

8401dd5

Signed-off-by: William Juan <[email protected]>

wjuan-AFK dismissed oschaaf’s stale review via 8401dd5 November 3, 2020 17:08

dubious90 reviewed Nov 3, 2020

View reviewed changes

dubious90 assigned mum4k Nov 3, 2020

dubious90 added waiting-for-changes A PR waiting for comments to be resolved and changes to be applied. and removed waiting-for-review A PR waiting for a review. labels Nov 3, 2020

Formatting and commenting changes for cl feedback. Using raw string.

8ee66df

Signed-off-by: William Juan <[email protected]>

wjuan-AFK added waiting-for-review A PR waiting for a review. and removed waiting-for-changes A PR waiting for comments to be resolved and changes to be applied. labels Nov 5, 2020

mum4k approved these changes Nov 5, 2020

View reviewed changes

mum4k assigned dubious90 and unassigned mum4k Nov 5, 2020

dubious90 approved these changes Nov 5, 2020

View reviewed changes

dubious90 merged commit 525c796 into envoyproxy:master Nov 5, 2020

wjuan-AFK deleted the CLIChangeForFileBasedRequestSource branch June 9, 2021 15:27

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Cli change for file based request source #563

Cli change for file based request source #563

wjuan-AFK commented Oct 22, 2020

repokitteh-read-only bot commented Nov 3, 2020

wjuan-AFK commented Nov 3, 2020

repokitteh-read-only bot commented Nov 3, 2020

oschaaf left a comment

oschaaf Nov 3, 2020

wjuan-AFK Nov 3, 2020

wjuan-AFK Nov 3, 2020

wjuan-AFK commented Nov 3, 2020

repokitteh-read-only bot commented Nov 3, 2020

dubious90 left a comment

dubious90 Nov 3, 2020

wjuan-AFK Nov 5, 2020

dubious90 Nov 3, 2020

dubious90 Nov 3, 2020

wjuan-AFK Nov 5, 2020

dubious90 Nov 3, 2020

wjuan-AFK Nov 5, 2020

wjuan-AFK commented Nov 5, 2020

repokitteh-read-only bot commented Nov 5, 2020

mum4k left a comment

mum4k commented Nov 5, 2020

Cli change for file based request source #563

Cli change for file based request source #563

Conversation

wjuan-AFK commented Oct 22, 2020

repokitteh-read-only bot commented Nov 3, 2020

wjuan-AFK commented Nov 3, 2020

repokitteh-read-only bot commented Nov 3, 2020

oschaaf left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wjuan-AFK commented Nov 3, 2020

repokitteh-read-only bot commented Nov 3, 2020

dubious90 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wjuan-AFK commented Nov 5, 2020

repokitteh-read-only bot commented Nov 5, 2020

mum4k left a comment

Choose a reason for hiding this comment

mum4k commented Nov 5, 2020