This is an automated email from the ASF dual-hosted git repository.
xuetaoli pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/dubbo-go-pixiu-samples.git
The following commit(s) were added to refs/heads/main by this push:
new 2f0ecf9 fix: fix CI problems (#129)
2f0ecf9 is described below
commit 2f0ecf9e712a54f01ffbf27f70621fef16e4a5dc
Author: 陈乐樂 <[email protected]>
AuthorDate: Mon Mar 23 09:41:26 2026 +0800
fix: fix CI problems (#129)
* feat: add kvcache samples
* fix some problems
* fix:fix some bugs and add headers
* feat:add integrate test
* feat: add descriptions to readme
* fix: fix ci problems
* remove replace
* replace gomod gosum
---
ai/kvcache/mock/pixiu/conf.yaml | 10 ++++++++--
ai/kvcache/mock/server/app/main.go | 27 +++++++++++++++++++++++----
ai/kvcache/mock/server/engine-a/main.go | 16 ++++++++++++++++
ai/kvcache/mock/server/engine-b/main.go | 16 ++++++++++++++++
ai/kvcache/real-engine/pixiu/conf.yaml | 10 ++++++++--
go.sum | 2 +-
6 files changed, 72 insertions(+), 9 deletions(-)
diff --git a/ai/kvcache/mock/pixiu/conf.yaml b/ai/kvcache/mock/pixiu/conf.yaml
index 7c50c85..313ad97 100644
--- a/ai/kvcache/mock/pixiu/conf.yaml
+++ b/ai/kvcache/mock/pixiu/conf.yaml
@@ -78,14 +78,20 @@ static_resources:
- name: "mock_llm"
lb_policy: "round_robin"
endpoints:
- - id: "mock-llm-a"
+ - ID: "mock-llm-a"
socket_address:
address: "127.0.0.1"
port: 18091
- - id: "mock-llm-b"
+ llm_meta:
+ retry_policy:
+ name: "NoRetry"
+ - ID: "mock-llm-b"
socket_address:
address: "127.0.0.1"
port: 18092
+ llm_meta:
+ retry_policy:
+ name: "NoRetry"
shutdown_config:
timeout: "10s"
diff --git a/ai/kvcache/mock/server/app/main.go
b/ai/kvcache/mock/server/app/main.go
index 957201f..95a676d 100644
--- a/ai/kvcache/mock/server/app/main.go
+++ b/ai/kvcache/mock/server/app/main.go
@@ -105,10 +105,11 @@ func main() {
preferredEndpoint := envOrDefault("PREFERRED_ENDPOINT_ID", "mock-llm-b")
engineAID := envOrDefault("LLM_A_ID", "mock-llm-a")
engineBID := envOrDefault("LLM_B_ID", "mock-llm-b")
+ responseDelay := envDurationMSOrDefault("MOCK_LLM_RESPONSE_DELAY_MS",
150)
controller := buildControllerMux(preferredEndpoint)
- engineA := buildEngineAMux(engineAID)
- engineB := buildEngineBMux(engineBID)
+ engineA := buildEngineAMux(engineAID, responseDelay)
+ engineB := buildEngineBMux(engineBID, responseDelay)
errCh := make(chan error, 3)
go serve("mock-controller", controllerAddr, controller, errCh)
@@ -239,7 +240,7 @@ func handleTokenEvent(stats *controllerStats, w
http.ResponseWriter, r *http.Req
writeJSON(w, http.StatusOK, eventResp{EventID: nextEventID(op),
NumTokens: len(req.Tokens)})
}
-func buildEngineAMux(engineID string) http.Handler {
+func buildEngineAMux(engineID string, responseDelay time.Duration)
http.Handler {
stats := &engineAStats{}
mux := http.NewServeMux()
@@ -294,6 +295,9 @@ func buildEngineAMux(engineID string) http.Handler {
stats.mu.Lock()
stats.chatCalls++
stats.mu.Unlock()
+ if responseDelay > 0 {
+ time.Sleep(responseDelay)
+ }
resp := llmResponse{
ID: nextEventID("chatcmpl"),
@@ -315,7 +319,7 @@ func buildEngineAMux(engineID string) http.Handler {
return mux
}
-func buildEngineBMux(engineID string) http.Handler {
+func buildEngineBMux(engineID string, responseDelay time.Duration)
http.Handler {
stats := &engineBStats{}
mux := http.NewServeMux()
@@ -358,6 +362,9 @@ func buildEngineBMux(engineID string) http.Handler {
stats.mu.Lock()
stats.chatCalls++
stats.mu.Unlock()
+ if responseDelay > 0 {
+ time.Sleep(responseDelay)
+ }
resp := llmResponse{
ID: nextEventID("chatcmpl"),
@@ -457,6 +464,18 @@ func envOrDefault(key string, fallback string) string {
return val
}
+func envDurationMSOrDefault(key string, fallbackMS int) time.Duration {
+ val, ok := os.LookupEnv(key)
+ if !ok || strings.TrimSpace(val) == "" {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ ms, err := strconv.Atoi(strings.TrimSpace(val))
+ if err != nil || ms < 0 {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ return time.Duration(ms) * time.Millisecond
+}
+
func max(a int, b int) int {
if a > b {
return a
diff --git a/ai/kvcache/mock/server/engine-a/main.go
b/ai/kvcache/mock/server/engine-a/main.go
index bb24e05..cdaa6d0 100644
--- a/ai/kvcache/mock/server/engine-a/main.go
+++ b/ai/kvcache/mock/server/engine-a/main.go
@@ -71,6 +71,7 @@ var eventCounter uint64
func main() {
addr := envOrDefault("LLM_A_ADDR", ":18091")
engineID := envOrDefault("LLM_A_ID", "mock-llm-a")
+ responseDelay := envDurationMSOrDefault("MOCK_LLM_RESPONSE_DELAY_MS",
150)
stats := &engineStats{}
mux := http.NewServeMux()
@@ -127,6 +128,9 @@ func main() {
stats.mu.Lock()
stats.chatCalls++
stats.mu.Unlock()
+ if responseDelay > 0 {
+ time.Sleep(responseDelay)
+ }
resp := llmResponse{
ID: nextEventID("chatcmpl"),
@@ -229,3 +233,15 @@ func envOrDefault(key string, fallback string) string {
}
return val
}
+
+func envDurationMSOrDefault(key string, fallbackMS int) time.Duration {
+ val, ok := os.LookupEnv(key)
+ if !ok || strings.TrimSpace(val) == "" {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ ms, err := strconv.Atoi(strings.TrimSpace(val))
+ if err != nil || ms < 0 {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ return time.Duration(ms) * time.Millisecond
+}
diff --git a/ai/kvcache/mock/server/engine-b/main.go
b/ai/kvcache/mock/server/engine-b/main.go
index c5ebcab..caa1e92 100644
--- a/ai/kvcache/mock/server/engine-b/main.go
+++ b/ai/kvcache/mock/server/engine-b/main.go
@@ -64,6 +64,7 @@ var eventCounter uint64
func main() {
addr := envOrDefault("LLM_B_ADDR", ":18092")
engineID := envOrDefault("LLM_B_ID", "mock-llm-b")
+ responseDelay := envDurationMSOrDefault("MOCK_LLM_RESPONSE_DELAY_MS",
150)
stats := &engineStats{}
mux := http.NewServeMux()
@@ -107,6 +108,9 @@ func main() {
stats.mu.Lock()
stats.chatCalls++
stats.mu.Unlock()
+ if responseDelay > 0 {
+ time.Sleep(responseDelay)
+ }
resp := llmResponse{
ID: nextEventID("chatcmpl"),
@@ -150,3 +154,15 @@ func envOrDefault(key string, fallback string) string {
}
return val
}
+
+func envDurationMSOrDefault(key string, fallbackMS int) time.Duration {
+ val, ok := os.LookupEnv(key)
+ if !ok || strings.TrimSpace(val) == "" {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ ms, err := strconv.Atoi(strings.TrimSpace(val))
+ if err != nil || ms < 0 {
+ return time.Duration(fallbackMS) * time.Millisecond
+ }
+ return time.Duration(ms) * time.Millisecond
+}
diff --git a/ai/kvcache/real-engine/pixiu/conf.yaml
b/ai/kvcache/real-engine/pixiu/conf.yaml
index 4c5e56c..2d47865 100644
--- a/ai/kvcache/real-engine/pixiu/conf.yaml
+++ b/ai/kvcache/real-engine/pixiu/conf.yaml
@@ -82,14 +82,20 @@ static_resources:
- name: "real_llm"
lb_policy: "round_robin"
endpoints:
- - id: "${ENGINE_ENDPOINT_A_ID}"
+ - ID: "${ENGINE_ENDPOINT_A_ID}"
socket_address:
address: "${ENGINE_ENDPOINT_A_HOST}"
port: ${ENGINE_ENDPOINT_A_PORT}
- - id: "${ENGINE_ENDPOINT_B_ID}"
+ llm_meta:
+ retry_policy:
+ name: "NoRetry"
+ - ID: "${ENGINE_ENDPOINT_B_ID}"
socket_address:
address: "${ENGINE_ENDPOINT_B_HOST}"
port: ${ENGINE_ENDPOINT_B_PORT}
+ llm_meta:
+ retry_policy:
+ name: "NoRetry"
shutdown_config:
timeout: "30s"
diff --git a/go.sum b/go.sum
index 96ccaa0..4fbaf84 100644
--- a/go.sum
+++ b/go.sum
@@ -2175,4 +2175,4 @@ rsc.io/sampler v1.3.0/go.mod
h1:T1hPZKmBbMNahiBKFy5HrXp6adAjACjK9JXDnKaTXpA=
sigs.k8s.io/yaml v1.1.0/go.mod h1:UJmg0vDUVViEyp3mgSv9WPwZCDxu4rQW1olrI1uml+o=
sigs.k8s.io/yaml v1.2.0 h1:kr/MCeFWJWTwyaHoR9c8EjH9OumOmoF9YGiZd7lFm/Q=
sigs.k8s.io/yaml v1.2.0/go.mod h1:yfXDCHCao9+ENCvLSE62v9VSji2MKu5jeNfTrofGhJc=
-sourcegraph.com/sourcegraph/appdash v0.0.0-20190731080439-ebfcffb1b5c0/go.mod
h1:hI742Nqp5OhwiqlzhgfbWU4mW4yO10fP+LoT9WOswdU=
+sourcegraph.com/sourcegraph/appdash v0.0.0-20190731080439-ebfcffb1b5c0/go.mod
h1:hI742Nqp5OhwiqlzhgfbWU4mW4yO10fP+LoT9WOswdU=
\ No newline at end of file
