fix: don't panic for some errors in modelstate #42
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
Signed-off-by: Travis Johnson <[email protected]>
Signed-off-by: Travis Johnson <[email protected]>
ckadner
requested review from
njhill
and removed request for
pvaneck and
chinhuang007
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: njhill, tjohnson31415 The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
/lgtm |
Jooho
referenced
this pull request
in Jooho/modelmesh-runtime-adapter
May 1, 2023
… - bug fixes Signed-off-by: jooho <[email protected]>
Jooho
referenced
this pull request
in red-hat-data-services/modelmesh-runtime-adapter
May 16, 2023
#### Motivation
If a model load timeout is hit a panic is generated in the modelStateManager:
```
panic: interface conversion: interface {} is nil, not *mmesh.LoadModelResponse
goroutine 1397 [running]:
github.com/kserve/modelmesh-runtime-adapter/model-serving-puller/server.(*modelStateManager).loadModel(...)
/opt/app-root/src/model-serving-puller/server/modelstate.go:107
github.com/kserve/modelmesh-runtime-adapter/model-serving-puller/server.(*PullerServer).LoadModel(0xc00016c550, {0x12df4f8, 0xc0005fc960}, 0x485800)
/opt/app-root/src/model-serving-puller/server/server.go:116 +0xad
github.com/kserve/modelmesh-runtime-adapter/internal/proto/mmesh._ModelRuntime_LoadModel_Handler({0x10219e0, 0xc00016c550}, {0x12df4f8, 0xc0005fc960}, 0xc0004602a0, 0x0)
/opt/app-root/src/internal/proto/mmesh/model-runtime_grpc.pb.go:181 +0x170
google.golang.org/grpc.(*Server).processUnaryRPC(0xc00054a1c0, {0x12f37d0, 0xc0006ca4e0}, 0xc000b326c0, 0xc00027daa0, 0x1a6f780, 0x0)
/remote-source/deps/gomod/pkg/mod/google.golang.org/[email protected]/server.go:1301 +0xb03
google.golang.org/grpc.(*Server).handleStream(0xc00054a1c0, {0x12f37d0, 0xc0006ca4e0}, 0xc000b326c0, 0x0)
/remote-source/deps/gomod/pkg/mod/google.golang.org/[email protected]/server.go:1642 +0xa2a
google.golang.org/grpc.(*Server).serveStreams.func1.2()
/remote-source/deps/gomod/pkg/mod/google.golang.org/[email protected]/server.go:938 +0x98
created by google.golang.org/grpc.(*Server).serveStreams.func1
/remote-source/deps/gomod/pkg/mod/google.golang.org/[email protected]/server.go:936 +0x294
```
In a couple of error cases in `submitRequest`, `nil` is returned as the first return value with the error. The code in `loadModel` and `unloadModel` always attempts to cast the value to a pointer to a response, but this will panic if attempting to convert `nil`.
#### Modifications
- add a test to reproduce the panic
- change the code to use a comma-ok type assertion instead of panicking
#### Result
The puller/adapter doesn't crash when a model load times out.
Signed-off-by: Travis Johnson <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Motivation
If a model load timeout is hit a panic is generated in the modelStateManager:
In a couple of error cases in
submitRequest,nilis returned as the first return value with the error. The code inloadModelandunloadModelalways attempts to cast the value to a pointer to a response, but this will panic if attempting to convertnil.Modifications
Result
The puller/adapter doesn't crash when a model load times out.