[v2] Implement sanitizers to operate on OTLP data #5551

varshith257 · 2024-06-10T07:45:48Z

Which problem is this PR solving?

Part of Transform v1 sanitizers into processors or something else #5545

Description of the changes

Rewritten sanitizers for v2 with the existing sanitizers to achieve parity with v1

How was this change tested?

Not Yet

Checklist

I have read https://github.com/jaegertracing/jaeger/blob/master/CONTRIBUTING_GUIDELINES.md
I have signed all commits
I have added unit tests for the new functionality
I have run lint and test steps successfully
- for jaeger: make lint test
- for jaeger-ui: yarn lint and yarn test

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

codecov · 2024-06-10T07:52:54Z

Codecov Report

Attention: Patch coverage is 0% with 64 lines in your changes missing coverage. Please review.

Project coverage is 96.27%. Comparing base (65f6cb9) to head (1062ba0).

Files	Patch %	Lines
...orters/storageexporter/sanitizer/utf8_sanitizer.go	0.00%	40 Missing ⚠️
...exporter/sanitizer/empty_service_name_sanitizer.go	0.00%	13 Missing ⚠️
...l/exporters/storageexporter/sanitizer/sanitizer.go	0.00%	11 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5551      +/-   ##
==========================================
- Coverage   96.65%   96.27%   -0.38%     
==========================================
  Files         342      345       +3     
  Lines       16519    16583      +64     
==========================================
  Hits        15966    15966              
- Misses        363      427      +64     
  Partials      190      190

Flag	Coverage Δ
badger_v1	`8.05% <ø> (ø)`
badger_v2	`1.81% <ø> (ø)`
cassandra-3.x-v1	`16.61% <ø> (ø)`
cassandra-3.x-v2	`1.74% <ø> (ø)`
cassandra-4.x-v1	`16.61% <ø> (ø)`
cassandra-4.x-v2	`1.74% <ø> (ø)`
elasticsearch-6.x-v1	`18.78% <ø> (ø)`
elasticsearch-7.x-v1	`18.84% <ø> (ø)`
elasticsearch-8.x-v1	`19.02% <ø> (ø)`
elasticsearch-8.x-v2	`1.81% <ø> (+0.01%)`	⬆️
grpc_v1	`9.52% <ø> (ø)`
grpc_v2	`7.14% <ø> (+0.01%)`	⬆️
kafka	`9.74% <ø> (ø)`
memory_v2	`1.81% <ø> (ø)`
opensearch-1.x-v1	`18.89% <ø> (+0.01%)`	⬆️
opensearch-2.x-v1	`18.89% <ø> (ø)`
opensearch-2.x-v2	`1.81% <ø> (ø)`
unittests	`94.71% <0.00%> (-0.37%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

varshith257 · 2024-06-10T07:53:18Z

@yurishkuro @albertteoh PTAL. I have rewritten the initial configuration of existing sanitizers to operate on OTLP data format. I have implemented them based on my understanding of OTLP data. Correct me, if I am wrong with understanding of OTLP data

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

cmd/collector/app/sanitizer_v2/service_name_sanitizer.go

cmd/collector/app/sanitizer_v2/sanitizer.go

yurishkuro · 2024-06-10T15:36:41Z

cmd/collector/app/sanitizer_v2/empty_service_name_sanitizer.go

+// sanitizeEmptyServiceName sanitizes the service names in the span attributes.
+func sanitizeEmptyServiceName(span ptrace.Span) ptrace.Span {
+ attributes := span.Attributes()
+ serviceNameAttr, ok := attributes.Get("service.name")


service name is stored in the Resource attributes, not in the span.

@yurishkuro span.Resource undefined (type ptrace.Span has no field or method Resource)compilerMissingFieldOrMethod

This means we need to access the resource from the ptrace.ResourceSpans. I think we need to work with ptrace.Traces instead of individual ptrace.Span and access the ResourceSpans which contains the Resource and associated ScopeSpans. Then we need to iterate through each ScopeSpans and their Spans. Am I in the right direction?

I think we need to work with ptrace.Traces

@james-ryans this is an interesting situation. In v1 pipeline all non-span data (like batch.Process) was first denormalized into each span so that the unit of work in the pipeline was a single span. In v2 storage API we want to support batching for better efficiency, but it would also mean that unless storage supports storing normalized data (like storing resource / scope records separately, which is unlikely in k/v stores world), each storage would need to do its own denormalization. It also creates challenges to shared capabilities like sanitizers here, since they need to operate on data uniformly before the storage is called.

We can provide a shared lib to do the flattening / denormalization of resources[]/scopes[]/spans[] into just spans[] (essentially otlp2jaeger does that today already), but I don't think it's a good solution considering that OTLP spans cannot distinguish between resource/scope/span attributes without some magic prefixes / naming. In v1 model we had a clear slot Span.Process to maintain that distinction. I am thinking that in v2 storage implementations their internal data model should also be Span{ Resource=..., Scope=..., rest of span }, to avoid unnecessary flattening.

Coming back to sanitizers, I think Sanitize(pdata.Traces) is a good starting point signature. Some sanitizers would only look at Resources in that input, while others might look at everything. The alternative would be to have distinct sanitizers for Resource, Scope, and Span, but I don't think we need this complexity at this point.

@varshith257 does this answer your question?

does this answer your question?

Yes. Thanks for it

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 · 2024-06-11T14:50:52Z

@yurishkuro PTAL and if it's ok, I will move ahead to add tests for them

cmd/collector/app/sanitizer_v2/sanitizer.go

cmd/collector/app/sanitizer_v2/service_name_sanitizer.go

cmd/collector/app/sanitizer_v2/utf8_sanitizer.go

cmd/collector/app/sanitizer_v2/empty_service_name_sanitizer.go

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 · 2024-06-12T02:48:44Z

@yurishkuro PTAL

cmd/jaeger/internal/exporters/storageexporter/sanitizer/sanitizer.go

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

…zer.go Co-authored-by: Yuri Shkuro <yurishkuro@users.noreply.github.com> Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

yurishkuro · 2024-06-12T03:07:09Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

+)
+
+// NewUTF8Sanitizer creates a UTF8 sanitizer.
+func NewUTF8Sanitizer() SanitizeTraces {


please add a test

yurishkuro · 2024-06-12T03:07:23Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/sanitizer.go

+type SanitizeTraces func(traces ptrace.Traces) ptrace.Traces
+
+// NewStandardSanitizers are automatically applied by SpanProcessor.
+func NewStandardSanitizers() []SanitizeTraces {


you need to call this from the exporter

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

cmd/jaeger/internal/exporters/storageexporter/sanitizer/empty_service_name_sanitizer.go

yurishkuro · 2024-06-12T03:13:51Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/empty_service_name_sanitizer.go

+ for i := 0; i < resourceSpans.Len(); i++ {
+ resourceSpan := resourceSpans.At(i)
+ attributes := resourceSpan.Resource().Attributes()
+ serviceNameAttr, ok := attributes.Get("service.name")


"service.name"

better to import semantic conventions and use a constant from there

yurishkuro · 2024-06-12T03:15:13Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

+ if !utf8.ValidString(k) {
+ originalKey := k
+ attributes.PutStr(invalidTagKey, k)
+ byteSlice := attributes.PutEmptyBytes(badUTF8Prefix + originalKey)


what happens to attributes.Range if you add attributes in the lambda function?

Modifying the map while iterating over it can lead to undefined behavior. Better, we should first collect the keys and values that need to be sanitized and then apply these changes after the iteration is complete.

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

…into v2-sanitizer

Co-authored-by: Yuri Shkuro <yurishkuro@users.noreply.github.com> Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

yurishkuro · 2024-06-12T22:36:18Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

+ for i := 0; i < resourceSpans.Len(); i++ {
+ resourceSpan := resourceSpans.At(i)
+
+ // Sanitize resource attributes


comments like this provide no value

yurishkuro · 2024-06-12T22:40:25Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

+ for _, k := range invalidKeys {
+ originalKey := k
+ attributes.PutStr(invalidTagKey, k)
+ binaryAttr := attributes.PutEmptyBytes(badUTF8Prefix + originalKey)


if the originalKey is a bad UTF8 string, you are still creating a new key with the same bad string

yurishkuro · 2024-06-12T22:45:24Z

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go

+
+ for k, v := range invalidValues {
+ originalValue := v
+ attributes.PutStr(k, invalidTagKey)


what does this achieve? I suggest you write a specification in English in the comments to the sanitizer what exactly it's trying to do with attributes. And I think starting with the unit tests in this case would be much more useful, because you can see exactly how input and output data are related. For example, I expect the following to be true:

traces1 := inputData // assume it has bad UTF8 strings traces2 := sanitizeUTF8(traces1) assert traces1 != traces2 // definitely changed something traces3 := sanitizeUTF8(traces2) assert traces2 == traces3 // the 2nd pass is no-op, since no more invalid data is present

varshith257 · 2024-06-29T19:26:46Z

Will complete this PR very soon( I had exams, so i hadn't done progress on this)

Manoramsharma · 2024-07-21T07:30:58Z

Hi @varshith257 If you are not working on this can I take this from here?

Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

intial configuration for v2 sanitizers

32db390

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

fix lint

7af6144

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 changed the title ~~[WIP] Implement v2 sanitizers~~ [v2] Implement sanitizers to operate on OTLP data Jun 10, 2024

fix lint

add2f2b

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

yurishkuro reviewed Jun 10, 2024

View reviewed changes

remove unrelated code

c0a30c3

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 requested a review from yurishkuro June 10, 2024 18:32

varshith257 and others added 2 commits June 11, 2024 08:24

updated sanitizers with ptrace.Traces

fd89e01

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

Merge branch 'main' into v2-sanitizer

c2841d8

varshith257 marked this pull request as ready for review June 11, 2024 03:03

varshith257 requested a review from a team as a code owner June 11, 2024 03:03

yurishkuro reviewed Jun 11, 2024

View reviewed changes

cmd/collector/app/sanitizer_v2/sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 11, 2024

View reviewed changes

cmd/collector/app/sanitizer_v2/service_name_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 11, 2024

View reviewed changes

cmd/collector/app/sanitizer_v2/utf8_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 11, 2024

View reviewed changes

cmd/collector/app/sanitizer_v2/empty_service_name_sanitizer.go Outdated Show resolved Hide resolved

varshith257 and others added 7 commits June 12, 2024 07:26

rename directory

53ae6bd

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

remove service-name-sanitizer in v2

0ef93f3

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

add compatibility section

7997c1b

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

add badUTF8 prefix

48a5229

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

Merge branch 'main' into v2-sanitizer

d029e7b

fix lint

cdd2633

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

fix

2c6fe5c

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 requested a review from yurishkuro June 12, 2024 02:48

yurishkuro added v2 changelog:exprimental Change to an experimental part of the code labels Jun 12, 2024

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go Outdated Show resolved Hide resolved

Update cmd/jaeger/internal/exporters/storageexporter/sanitizer/saniti…

7f0e9cc

…zer.go Co-authored-by: Yuri Shkuro <yurishkuro@users.noreply.github.com> Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/utf8_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/empty_service_name_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 12, 2024

View reviewed changes

cmd/jaeger/internal/exporters/storageexporter/sanitizer/empty_service_name_sanitizer.go Outdated Show resolved Hide resolved

yurishkuro reviewed Jun 12, 2024

View reviewed changes

varshith257 and others added 5 commits June 12, 2024 21:50

updated utf8 sanitizer

1d46f5d

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

Merge branch 'v2-sanitizer' of https://github.com/varshith257/jaeger …

45ee0aa

…into v2-sanitizer

rename vars

3464d25

Co-authored-by: Yuri Shkuro <yurishkuro@users.noreply.github.com> Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

fix

786a279

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

fix vars

28def9c

Signed-off-by: Vamshi Maskuri <gwcchintu@gmail.com>

varshith257 requested a review from yurishkuro June 12, 2024 17:11

yurishkuro reviewed Jun 12, 2024

View reviewed changes

varshith257 closed this Jun 13, 2024

varshith257 reopened this Jun 29, 2024

Merge branch 'main' into v2-sanitizer

75026c0

varshith257 added 2 commits July 21, 2024 13:03

Merge branch 'main' into v2-sanitizer

dab72f9

Merge branch 'main' into v2-sanitizer

1062ba0

Signed-off-by: Vamshi Maskuri <117595548+varshith257@users.noreply.github.com>

yurishkuro mentioned this pull request Oct 11, 2024

Transform v1 sanitizers into processors or something else #5545

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[v2] Implement sanitizers to operate on OTLP data #5551

[v2] Implement sanitizers to operate on OTLP data #5551

varshith257 commented Jun 10, 2024 •

edited

Loading

codecov bot commented Jun 10, 2024 •

edited

Loading

varshith257 commented Jun 10, 2024 •

edited

Loading

yurishkuro Jun 10, 2024

varshith257 Jun 10, 2024 •

edited

Loading

yurishkuro Jun 10, 2024

varshith257 Jun 11, 2024

varshith257 commented Jun 11, 2024

varshith257 commented Jun 12, 2024

yurishkuro Jun 12, 2024

yurishkuro Jun 12, 2024

yurishkuro Jun 12, 2024

yurishkuro Jun 12, 2024

varshith257 Jun 12, 2024

yurishkuro Jun 12, 2024

yurishkuro Jun 12, 2024

yurishkuro Jun 12, 2024

varshith257 commented Jun 29, 2024 •

edited

Loading

Manoramsharma commented Jul 21, 2024

[v2] Implement sanitizers to operate on OTLP data #5551

Are you sure you want to change the base?

[v2] Implement sanitizers to operate on OTLP data #5551

Conversation

varshith257 commented Jun 10, 2024 • edited Loading

Which problem is this PR solving?

Description of the changes

How was this change tested?

Checklist

codecov bot commented Jun 10, 2024 • edited Loading

Codecov Report

varshith257 commented Jun 10, 2024 • edited Loading

Choose a reason for hiding this comment

varshith257 Jun 10, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

varshith257 commented Jun 11, 2024

varshith257 commented Jun 12, 2024

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

varshith257 commented Jun 29, 2024 • edited Loading

Manoramsharma commented Jul 21, 2024

varshith257 commented Jun 10, 2024 •

edited

Loading

codecov bot commented Jun 10, 2024 •

edited

Loading

varshith257 commented Jun 10, 2024 •

edited

Loading

varshith257 Jun 10, 2024 •

edited

Loading

varshith257 commented Jun 29, 2024 •

edited

Loading