allow caller context during reloads() to cancel (#19687)

canceled callers might linger around longer,
can potentially overwhelm the system. Instead
provider a caller context and canceled callers
don't hold on to them.

Bonus: we have no reason to cache errors, we should
never cache errors otherwise we can potentially have
quorum errors creeping in unexpectedly. We should
let the cache when invalidating hit the actual resources
instead.

cmd/data-usage.go

@@ -79,12 +79,12 @@ func loadPrefixUsageFromBackend(ctx context.Context, objAPI ObjectLayer, bucket
 	prefixUsageCache.InitOnce(30*time.Second,
 		// No need to fail upon Update() error, fallback to old value.
 		cachevalue.Opts{ReturnLastGood: true, NoWait: true},
-		func() (map[string]uint64, error) {
+		func(ctx context.Context) (map[string]uint64, error) {
 			m := make(map[string]uint64)
 			for _, pool := range z.serverPools {
 				for _, er := range pool.sets {
 					// Load bucket usage prefixes
-					ctx, done := context.WithTimeout(context.Background(), 2*time.Second)
+					ctx, done := context.WithTimeout(ctx, 2*time.Second)
 					ok := cache.load(ctx, er, bucket+slashSeparator+dataUsageCacheName) == nil
 					done()
 					if ok {
@@ -107,7 +107,7 @@ func loadPrefixUsageFromBackend(ctx context.Context, objAPI ObjectLayer, bucket
 		},
 	)
 
-	return prefixUsageCache.Get()
+	return prefixUsageCache.GetWithCtx(ctx)
 }
 
 func loadDataUsageFromBackend(ctx context.Context, objAPI ObjectLayer) (DataUsageInfo, error) {