Issue with metric refresh interval

[varungup90]

🐛 Describe the bug

I noticed that metrics are not refreshed correctly per the interval. In below logs, interval set is 1s, but for decode-0 pod, metric is refreshed twice for same interval.

I1029 23:28:23.046741       1 cache_metrics.go:453] "Updating model metric" pod="vllm-2p2d-tp2-1roleset-roleset-wk7x7-decode-5dcf575575-0" model="qwen3-32b" generation_token_total={"Value":1315505} avg_generation_throughput_toks_per_s={"Value":1217.4834983307785}
I1029 23:28:23.046754       1 cache_metrics.go:186] 


I1029 23:28:23.049670       1 gateway_rsp_body.go:157] request end, requestID: fead47f2-39d2-4bf7-ada8-a3e9c67a21d2 - targetPod: 192.168.0.102:8000, elapsed: 6.59526097s
I1029 23:28:23.089137       1 cache_metrics.go:453] "Updating model metric" pod="vllm-2p2d-tp2-1roleset-roleset-wk7x7-decode-5dcf575575-0" model="qwen3-32b" generation_token_total={"Value":1315565} avg_generation_throughput_toks_per_s={"Value":1415.1147987115946}
I1029 23:28:23.089151       1 cache_metrics.go:186] 


I1029 23:28:23.089311       1 cache_metrics.go:453] "Updating model metric" pod="vllm-2p2d-tp2-1roleset-roleset-wk7x7-decode-5dcf575575-1" model="qwen3-32b" generation_token_total={"Value":1347888} avg_generation_throughput_toks_per_s={"Value":1370.1406205634362}

Steps to Reproduce

I added a log message in cache_metrics.go to print the values and run benchmark test.

Expected behavior

there should be only one record for each pod for every metric in the interval

Environment

NA

[googs1025]

Could this be a problem with duplicate pods in metaPods? 🤔