Envoy sidecar memory leak in Istio 1.20+ — anyone else seeing RSS growth over 72h?

After upgrading to Istio 1.20, we're seeing Envoy sidecars grow from ~200MB to ~1.2GB RSS over 72 hours. No OOM kills yet (limits at 1.5GB) but it's eating into node memory. We've ruled out application-level leaks — the sidecar itself is growing. Prometheus metrics show stable connection counts and request rates. Has anyone else hit this? Did you find a config knob (maybe outlier detection or connection pooling) that triggers it, or is it a known upstream issue? Jurisdiction: INTL