jiazhizhong/higress
1
0
Fork 0
mirror of https://github.com/alibaba/higress.git synced 2026-06-09 20:57:32 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat: implement apiToken failover mechanism (#1256)

Browse Source
This commit is contained in:
Se7en
2024-11-16 19:03:09 +08:00
committed by GitHub
parent f2a5df3949
commit d24123a55f
33 changed files with 1558 additions and 1231 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
plugins/wasm-go/extensions/ai-proxy/provider/ai360.go
View File

@@ -1,14 +1,12 @@
package provider
import (
"encoding/json"
"errors"
"fmt"
"github.com/higress-group/proxy-wasm-go-sdk/proxywasm"
"github.com/higress-group/proxy-wasm-go-sdk/proxywasm/types"
"net/http"
"github.com/alibaba/higress/plugins/wasm-go/extensions/ai-proxy/util"
"github.com/alibaba/higress/plugins/wasm-go/pkg/wrapper"
"github.com/higress-group/proxy-wasm-go-sdk/proxywasm/types"
)
// ai360Provider is the provider for 360 OpenAI service.
@@ -46,10 +44,7 @@ func (m *ai360Provider) OnRequestHeaders(ctx wrapper.HttpContext, apiName ApiNam
if apiName != ApiNameChatCompletion && apiName != ApiNameEmbeddings {
return types.ActionContinue, errUnsupportedApiName
}
_ = util.OverwriteRequestHost(ai360Domain)
_ = proxywasm.RemoveHttpRequestHeader("Accept-Encoding")
_ = proxywasm.RemoveHttpRequestHeader("Content-Length")
_ = proxywasm.ReplaceHttpRequestHeader("Authorization", m.config.GetRandomToken())
m.config.handleRequestHeaders(m, ctx, apiName, log)
// Delay the header processing to allow changing streaming mode in OnRequestBody
return types.HeaderStopIteration, nil
}
@@ -58,47 +53,12 @@ func (m *ai360Provider) OnRequestBody(ctx wrapper.HttpContext, apiName ApiName,
if apiName != ApiNameChatCompletion && apiName != ApiNameEmbeddings {
return types.ActionContinue, errUnsupportedApiName
}
if apiName == ApiNameChatCompletion {
return m.onChatCompletionRequestBody(ctx, body, log)
}
if apiName == ApiNameEmbeddings {
return m.onEmbeddingsRequestBody(ctx, body, log)
}
return types.ActionContinue, errUnsupportedApiName
return m.config.handleRequestBody(m, m.contextCache, ctx, apiName, body, log)
}
func (m *ai360Provider) onChatCompletionRequestBody(ctx wrapper.HttpContext, body []byte, log wrapper.Log) (types.Action, error) {
request := &chatCompletionRequest{}
if err := decodeChatCompletionRequest(body, request); err != nil {
return types.ActionContinue, err
}
if request.Model == "" {
return types.ActionContinue, errors.New("missing model in chat completion request")
}
// 映射模型
mappedModel := getMappedModel(request.Model, m.config.modelMapping, log)
if mappedModel == "" {
return types.ActionContinue, errors.New("model becomes empty after applying the configured mapping")
}
ctx.SetContext(ctxKeyFinalRequestModel, mappedModel)
request.Model = mappedModel
return types.ActionContinue, replaceJsonRequestBody(request, log)
}
func (m *ai360Provider) onEmbeddingsRequestBody(ctx wrapper.HttpContext, body []byte, log wrapper.Log) (types.Action, error) {
request := &embeddingsRequest{}
if err := json.Unmarshal(body, request); err != nil {
return types.ActionContinue, fmt.Errorf("unable to unmarshal request: %v", err)
}
if request.Model == "" {
return types.ActionContinue, errors.New("missing model in embeddings request")
}
// 映射模型
mappedModel := getMappedModel(request.Model, m.config.modelMapping, log)
if mappedModel == "" {
return types.ActionContinue, errors.New("model becomes empty after applying the configured mapping")
}
ctx.SetContext(ctxKeyFinalRequestModel, mappedModel)
request.Model = mappedModel
return types.ActionContinue, replaceJsonRequestBody(request, log)
func (m *ai360Provider) TransformRequestHeaders(ctx wrapper.HttpContext, apiName ApiName, headers http.Header, log wrapper.Log) {
util.OverwriteRequestHostHeader(headers, ai360Domain)
util.OverwriteRequestAuthorizationHeader(headers, "Authorization "+m.config.GetApiTokenInUse(ctx))
headers.Del("Accept-Encoding")
headers.Del("Content-Length")
}