Files
Ilker ManapandClaude Opus 5.5 aa531b1053 ComfyUI models of one environment share one replica
flux and sf3d on one card handed it to each other, each starting ComfyUI
and loading its weights from nothing (BACKLOG H3, placement side). An
environment of two or more models now gets a hidden catalog entry,
comfyui-env-<name>: on its first member's profiles with the most memory
any member needs, on demand for the longest idle timeout, carrying every
member's weights, repositories and the environment's custom nodes. The
members get no replicas of their own; asking for one wants the entry.

The agent fetches each member's weights; the gate links all of them,
refusing two weights files of one name, links each member's Hugging Face
repository, and answers /v1/models for every member. The control plane
offers the replica under the members' names with each one's workflow and
checkpoint, which the router sends by the model asked for. A member's
sibling never evicts the replica serving it, and the members' load counts
as the entry's.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-14 13:55:34 +02:00

193 lines
6.4 KiB
Go

package comfy
import (
"encoding/json"
"fmt"
"io"
"net/http"
"net/http/httptest"
"net/url"
"os"
"os/exec"
"path/filepath"
"strings"
"sync"
"testing"
"time"
)
func TestTheBookReadsComfyUIsWebsocket(t *testing.T) {
b := newProgressBook()
for _, m := range []string{
`{"type":"status","data":{"status":{"exec_info":{"queue_remaining":1}}}}`,
`{"type":"execution_start","data":{"prompt_id":"p1"}}`,
`{"type":"executing","data":{"node":"5","prompt_id":"p1"}}`,
`{"type":"progress","data":{"value":12,"max":30,"prompt_id":"p1","node":"5"}}`,
} {
b.read([]byte(m))
}
p, _ := b.get("p1")
if p.State != "running" || p.Node != "5" || p.Step != 12 || p.Steps != 30 || p.Percent != 40 {
t.Errorf("%+v", p)
}
b.read([]byte(`{"type":"executing","data":{"node":null,"prompt_id":"p1"}}`))
if p, _ := b.get("p1"); p.State != "done" || p.Percent != 100 {
t.Errorf("finished: %+v", p)
}
b.read([]byte(`{"type":"execution_error","data":{"prompt_id":"p2","node_type":"KSampler","exception_message":"CUDA out of memory"}}`))
if p, _ := b.get("p2"); p.State != "failed" || p.Error != "KSampler: CUDA out of memory" {
t.Errorf("failed: %+v", p)
}
for i := range keptPrompts + 10 {
b.update(fmt.Sprint("x", i), func(*Progress) {})
}
if _, ok := b.get("p1"); ok || len(b.by) != keptPrompts {
t.Errorf("the book kept %d", len(b.by))
}
}
func TestTheGateHearsTheWorkflowsItLetsThrough(t *testing.T) {
var mu sync.Mutex
var sent map[string]any
comfy := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
if r.URL.Path == "/prompt" {
mu.Lock()
_ = json.NewDecoder(r.Body).Decode(&sent)
mu.Unlock()
fmt.Fprint(w, `{"prompt_id":"p9"}`)
}
}))
defer comfy.Close()
u, _ := url.Parse(comfy.URL)
book := newProgressBook()
gate := httptest.NewServer(newGate(u, "", []string{"m"}, book, "llmgw-gate-1"))
defer gate.Close()
resp, err := http.Post(gate.URL+"/prompt", "application/json", strings.NewReader(`{"prompt":{},"client_id":"router"}`))
if err != nil {
t.Fatal(err)
}
b, _ := io.ReadAll(resp.Body)
resp.Body.Close()
if !strings.Contains(string(b), "p9") {
t.Fatalf("the answer did not come through: %s", b)
}
mu.Lock()
if sent["client_id"] != "llmgw-gate-1" {
t.Errorf("ComfyUI was sent client %v: its progress would go to nobody listening", sent["client_id"])
}
mu.Unlock()
book.read([]byte(`{"type":"progress","data":{"value":3,"max":4,"prompt_id":"p9","node":"2"}}`))
resp, _ = http.Get(gate.URL + "/progress/p9")
var p Progress
_ = json.NewDecoder(resp.Body).Decode(&p)
resp.Body.Close()
if p.State != "running" || p.Percent != 75 {
t.Errorf("/progress: %+v", p)
}
if resp, _ := http.Get(gate.URL + "/progress/unknown"); resp.StatusCode != http.StatusNotFound {
t.Errorf("an unknown workflow: %d", resp.StatusCode)
}
}
func TestAWaitingWorkflowSaysHowManyAreAheadAndThenHowFarAlong(t *testing.T) {
var mu sync.Mutex
stage := 0
srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
mu.Lock()
s := stage
mu.Unlock()
switch r.URL.Path {
case "/prompt":
fmt.Fprint(w, `{"prompt_id":"me"}`)
case "/queue":
if s == 0 {
fmt.Fprint(w, `{"queue_running":[[1,"a"]],"queue_pending":[[3,"me"],[2,"b"],[4,"c"]]}`)
} else {
fmt.Fprint(w, `{"queue_running":[[3,"me"]],"queue_pending":[]}`)
}
case "/progress/me":
fmt.Fprint(w, `{"state":"running","node":"5","step":6,"steps":12,"percent":50}`)
case "/history/me":
if s < 2 {
fmt.Fprint(w, `{}`)
return
}
fmt.Fprint(w, `{"me":{"status":{"status_str":"success","completed":true},"outputs":{"9":{"images":[{"filename":"a.png","type":"temp"}]}}}}`)
case "/view":
w.Header().Set("Content-Type", "image/png")
fmt.Fprint(w, "png")
}
}))
defer srv.Close()
var seen []Progress
c := &Client{Poll: time.Millisecond}
done := make(chan error, 1)
go func() {
_, err := c.GenerateWatch(t.Context(), srv.URL, "", map[string]any{"5": map[string]any{"class_type": "KSampler"}},
func(p Progress) {
mu.Lock()
seen = append(seen, p)
stage++
mu.Unlock()
})
done <- err
}()
if err := <-done; err != nil {
t.Fatal(err)
}
if len(seen) < 2 || seen[0].State != "queued" || seen[0].Ahead != 2 {
t.Fatalf("%+v: queued behind the one running and the one numbered before it, wanted first", seen)
}
if seen[1].State != "running" || seen[1].Percent != 50 || seen[1].NodeClass != "KSampler" {
t.Errorf("then %+v", seen[1])
}
}
func TestANodeIsPutAtItsCommit(t *testing.T) {
if _, err := exec.LookPath("git"); err != nil {
t.Skip("no git")
}
repo := t.TempDir()
run := func(args ...string) string {
cmd := exec.Command("git", args...)
cmd.Dir = repo
cmd.Env = append(os.Environ(), "GIT_AUTHOR_NAME=t", "GIT_AUTHOR_EMAIL=t@t", "GIT_COMMITTER_NAME=t", "GIT_COMMITTER_EMAIL=t@t")
b, err := cmd.CombinedOutput()
if err != nil {
t.Fatalf("git %v: %v %s", args, err, b)
}
return strings.TrimSpace(string(b))
}
run("init", "-q")
run("config", "uploadpack.allowReachableSHA1InWant", "true")
_ = os.WriteFile(filepath.Join(repo, "__init__.py"), []byte("v1\n"), 0o644)
run("add", ".")
run("commit", "-qm", "one")
first := run("rev-parse", "HEAD")
_ = os.WriteFile(filepath.Join(repo, "__init__.py"), []byte("v2\n"), 0o644)
run("commit", "-qam", "two")
state := t.TempDir()
site, err := installNodes(t.Context(), state, "python3", []Node{{Repo: "file://" + repo, Commit: first}}, "", func(string) {})
if err != nil {
t.Fatal(err)
}
b, _ := os.ReadFile(filepath.Join(state, "custom_nodes", filepath.Base(repo), "__init__.py"))
if string(b) != "v1\n" || site != "" {
t.Errorf("the node is %q, site %q: wanted the commit named, not the branch's head", b, site)
}
if _, err := installNodes(t.Context(), state, "python3", []Node{{Repo: "file://" + repo, Commit: first,
Requirements: []string{"trimesh"}}}, "", func(string) {}); err == nil || !strings.Contains(err.Error(), "no uv") {
t.Errorf("packages without a uv: %v", err)
}
}
func TestANodesOwnRangeDecidesThePackagesItNames(t *testing.T) {
freeze := []byte("numpy==2.3.1\ntransformers==5.19.0\nhuggingface-hub==1.33.0\nPillow==12.3.0\n")
got := string(withoutNamed(freeze, []string{"transformers>=4.50.3,<5", "Huggingface_Hub<1", "trimesh"}))
if strings.Contains(got, "transformers") || strings.Contains(got, "huggingface") ||
!strings.Contains(got, "numpy==2.3.1") || !strings.Contains(got, "Pillow==12.3.0") {
t.Errorf("constraints %q", got)
}
}