highperfocused/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-11-11 03:17:47 +01:00
Code Issues Packages Projects Releases Wiki Activity

test: harden scheduler tests (#12662)

Browse Source 
* test: harden scheduler tests

This removes reschedDelay which was stale code, and adds
a new configurable timeout for the waitForVRAMRecovery so
tests can now set the timeout to be very short to avoid the
scheduler getting stuck and hitting a test timeout.

* test: tune tests for partial loads

Give stress tests more time when the model is split between CPU/GPU
This commit is contained in:
Daniel Hiltgen
2025-10-17 08:56:44 -07:00
committed by GitHub
parent 270679932f
commit 68e04c7ff8
10 changed files with 195 additions and 143 deletions
Download Patch File Download Diff File Expand all files Collapse all files

54
server/routes_harmony_streaming_test.go
View File

@@ -268,15 +268,15 @@ func TestChatHarmonyParserStreamingRealtime(t *testing.T) {
s := Server{
sched: &Scheduler{
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
reschedDelay: 100 * time.Millisecond,
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
waitForRecovery: 100 * time.Millisecond,
loadFn: func(req *LlmRequest, _ *ggml.GGML, _ discover.GpuInfoList, _ bool) bool {
req.successCh <- &runnerRef{
llama: &mock,
@@ -419,15 +419,15 @@ func TestChatHarmonyParserStreamingSimple(t *testing.T) {
s := Server{
sched: &Scheduler{
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
reschedDelay: 100 * time.Millisecond,
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
waitForRecovery: 100 * time.Millisecond,
loadFn: func(req *LlmRequest, _ *ggml.GGML, _ discover.GpuInfoList, _ bool) bool {
req.successCh <- &runnerRef{
llama: &mock,
@@ -601,15 +601,15 @@ func TestChatHarmonyParserStreaming(t *testing.T) {
s := Server{
sched: &Scheduler{
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
reschedDelay: 250 * time.Millisecond,
pendingReqCh: make(chan *LlmRequest, 1),
finishedReqCh: make(chan *LlmRequest, 1),
expiredCh: make(chan *runnerRef, 1),
unloadedCh: make(chan any, 1),
loaded: make(map[string]*runnerRef),
newServerFn: newMockServer(&mock),
getGpuFn: getGpuFn,
getCpuFn: getCpuFn,
waitForRecovery: 250 * time.Millisecond,
loadFn: func(req *LlmRequest, _ *ggml.GGML, _ discover.GpuInfoList, _ bool) bool {
req.successCh <- &runnerRef{
llama: &mock,