This repository has been archived by the owner on Jan 12, 2022. It is now read-only.
looker: log series with synthetic time source coupled to wall time (part 1) #1310
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Batch rename from #1009 was going into the 'wrong' direction :). Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Instead of having one option using the unit `ms` and another option using the unit `ns` use the same for both series types. With that, common constructor logic can be deduplicated much easier. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Also, set walltime coupling options for both cases. With this commit, there should be no behavioral change for log series as well as metric series. tsc shows no error. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
already with slight adjustments, but incomplete also does not compile yet Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
And validation logic. This starts to look good. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Interface / stubs were moved to base class, rebuild previous logic by filling in lastSampleSecondsSinceEpoch() and leapForward() Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
I saw on a this PR that looker artifact collection is now broken because it terminates the calling script (as of errexit opt) before that can invoke the collection commands. change paradigm: use errexit in invoke-looker.sh and call via a new shell -- capture exit code of that process. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Breaking change, for sure. Has to to be done. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
cannot serialize bigint to json, so take this route and do runtime type validation Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Also, there seems to be a formatting update. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Important: factor 1000 Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
queryLoki(): change req timeout cfg: bump 'request' timeout parameter (covering all request phases) by a little more than factor two. Kind of unexpected to help. But if it does: interesting to learn about _why_. Most importantly, I'd like to understand based on log output how many requests are really made during the polling loop and how those fail that currently do not generate log output. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Can be re-used in children Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Initialized on base class now. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
And also adjust child classes here and there. This is almost complete but does not compile yet. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Now that the LogSeries class also supports the loose walltime coupling the two cases here can be treated the same way, using generateNextFragmentOrSkip() Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
Make it so that `minLagSeconds` comes with a guarantee (never make the actual lag smaller than that). This makes the throttling criterion easier to build and read. This removes the need for validation (`mtls` code removal). This required proper introduction of `fragmentTimeLeapSeconds`. All these are nice. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
The stringified options were getting a little long. This wasn't useful anymore. What's more useful is showing the unique properties of the time series -- which are given by the query string. Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
also reduce time constant (follow-up from previous commit where I bumped from 5 to 12 but I think this is better) Signed-off-by: Dr. Jan-Philip Gehrcke <jp@opstrace.com>
jgehrcke
changed the title
|
good state, merging now to land CI improvements. follow-up pr for looker upcoming. |
|
which is caused by #1226 but shows a skewed symptom because of |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
From: #982
That's what this PR is mainly about. It does not fully implement this 'feature' yet, but is just a significant milestone on the path to that state.
In addition to that, this PR: