项目文件夹

文件
Asim Aslam a8ea60120e agent: make model retries opt-in and harden retry backoff (#3021)
Follow-up to #3017. A Generate runs the whole tool-execution turn, so
auto-retrying it re-runs already-executed (possibly side-effecting) tool
calls. Default ModelMaxAttempts 3 -> 1: retries are now opt-in via
ModelRetry. The timeout stays as a safety net.

Also harden ai.GenerateWithRetry: always back off between retries
(exponential, capped at 30s, default 200ms if unset) so an opt-in retry
can't busy-loop the provider even with Backoff=0.

Verified: go build, go test -race ./agent/... ./ai/, golangci-lint.

Co-authored-by: Claude <noreply@anthropic.com>
2026-06-24 12:42:15 +01:00

126 行
2.9 KiB
Go

package ai
import (
"context"
"errors"
"fmt"
"strings"
"time"
)
// StatusCoder is implemented by provider errors that expose an HTTP-like status code.
type StatusCoder interface {
StatusCode() int
}
// RetryError is returned when Generate is retried and still fails.
type RetryError struct {
Attempts int
Err error
}
func (e *RetryError) Error() string {
if e == nil {
return ""
}
return fmt.Sprintf("ai generate failed after %d attempt(s): %v", e.Attempts, e.Err)
}
func (e *RetryError) Unwrap() error {
if e == nil {
return nil
}
return e.Err
}
// GeneratePolicy controls timeout and retry behavior for a model call.
type GeneratePolicy struct {
Timeout time.Duration
MaxAttempts int
Backoff time.Duration
}
// GenerateWithRetry calls m.Generate with per-attempt timeout and bounded retry.
func GenerateWithRetry(ctx context.Context, m Model, req *Request, policy GeneratePolicy, opts ...GenerateOption) (*Response, error) {
if policy.MaxAttempts <= 0 {
policy.MaxAttempts = 1
}
if m == nil {
return nil, errors.New("ai model is nil")
}
var last error
for attempt := 1; attempt <= policy.MaxAttempts; attempt++ {
if err := ctx.Err(); err != nil {
return nil, err
}
callCtx := ctx
cancel := func() {}
if policy.Timeout > 0 {
callCtx, cancel = context.WithTimeout(ctx, policy.Timeout)
}
resp, err := m.Generate(callCtx, req, opts...)
cancel()
if err == nil {
return resp, nil
}
last = err
// Caller cancellation/deadline always wins and is not retried.
if ctx.Err() != nil {
return nil, ctx.Err()
}
if attempt == policy.MaxAttempts || !IsTransientError(err) {
if attempt > 1 || IsTransientError(err) {
return nil, &RetryError{Attempts: attempt, Err: err}
}
return nil, err
}
// Always back off between retries — exponential and capped — so an
// opt-in retry can never become a tight loop hammering the provider,
// even if Backoff was left at zero.
backoff := policy.Backoff
if backoff <= 0 {
backoff = 200 * time.Millisecond
}
if shift := attempt - 1; shift > 0 {
backoff <<= shift
}
if backoff > 30*time.Second {
backoff = 30 * time.Second
}
t := time.NewTimer(backoff)
select {
case <-ctx.Done():
if !t.Stop() {
<-t.C
}
return nil, ctx.Err()
case <-t.C:
}
}
return nil, &RetryError{Attempts: policy.MaxAttempts, Err: last}
}
// IsTransientError reports whether err is worth retrying at the provider boundary.
func IsTransientError(err error) bool {
if err == nil {
return false
}
if errors.Is(err, context.Canceled) {
return false
}
if errors.Is(err, context.DeadlineExceeded) {
return true
}
var sc StatusCoder
if errors.As(err, &sc) {
code := sc.StatusCode()
return code == 429 || code >= 500
}
msg := strings.ToLower(err.Error())
return strings.Contains(msg, "rate limit") || strings.Contains(msg, "too many requests") || strings.Contains(msg, "timeout") || strings.Contains(msg, "temporar")
}