REQUEST: New membership for aeron-gh #163

GitHub Username

@aeron-gh

Membership Level Requested

Member

Requirements

• I have reviewed the community membership guidelines and understand the requirements for my requested membership level
• I have enabled 2FA on my GitHub account
• I have subscribed to the Volcano Mailing List
• I am actively contributing to 1 or more Volcano subprojects
• I have the required number of sponsors that meet the sponsor requirements for my requested role as listed in the community membership guidelines
• I have spoken to my sponsors ahead of this application, and they have agreed to sponsor my application
• I have verified that my sponsors meet the requirements for my requested membership level (e.g.,Member requires 2 approvers; Reviewer/Approver requires 2 maintainers) as specified in the community guidelines

Sponsor 1

@YaoZengzeng

Sponsor 2

@LiZhenCheng9527

List of contributions to the Volcano project

Contributing since June 2026. 48 pull requests authored across the org: 35 merged, 12 open, 1 closed. 33 issues filed across kthena, volcano, website and dashboard, 18 of them closed by my own merged PRs.

kthena, 26 merged (all authored / merged)

• #1431 runtime: use Redis SCAN instead of KEYS
• #1440 helm: document replicas values
• #1449 helm: drop unused webhook values
• #1463 router: skip KV cache ownership written before a pod restart
• #1522 cleanup: remove unused extractPodNameFromIdentifier
• #1548 cli: replace ModelInfer with ModelServing in README
• #1572 runtime: drop token ids and hash lists from kv cache logs
• #1575 cli: fix template directory path in README
• #1581 runtime: name the mapping key TTL
• #1586 docs: use the routable model name in inference examples
• #1590 logging: clean up remaining error log messages
• #1597 docs: match the real event handler log line in kvcache troubleshooting
• #1598 docs: sync quick-start booster asset with the applied example
• #1606 router: fail conf test when an expected error is missing
• #1621 docs: fix the router ConfigMap example key and nesting
• #1646 rbac: allow controller-manager to delete modelservers and modelroutes
• #1650 router: make the random score plugin safe for concurrent requests
• #1651 cli: honor KUBECONFIG when loading the kubeconfig
• #1662 helm: honor the controller manager webhook toggle
• #1703 runtime: key counter rename rules by the parsed family name
• #1719 router: fail the modelroute test cleanly when the cache wait loses
• #1721 e2e: pin the llm-d inference sim image
• #1769 router: report a closed queue instead of a client disconnect
• #1777 docs: correct the Kubernetes prerequisite to 1.28
• #1780 e2e: add a maxSurge rolling update test with partition
• #1798 datastore: emit the Pod delete event when DeleteModelServer evicts a pod

volcano, 9 merged (all authored / merged)

• #5487 fix(webhook): add separator between PodGroup validation errors
• #5538 fix(cli): correct flag descriptions in vcctl pod command
• #5593 fix(networkqos): correct wording and spacing in option flag descriptions
• #5610 fix(webhook): separate CronJob validation errors with "; "
• #5632 fix: remove duplicated words in vcctl and network-qos messages
• #5636 fix(cli): dedup vcctl pod list by namespace and name
• #5665 feat(cli): sort vcctl pod list output by namespace and name
• #5675 fix(jobflow): settle the running state to Failed when a job is Terminated
• #5743 fix(sharding): keep allocation-rate default range when arguments omitted

Open PRs: 8 in kthena, 2 in volcano, 2 in website. volcano #5679 is approved, and both website PRs have lgtm. The largest is kthena #1738, which makes the autoscaler honor PrometheusMetricSource.auth (bearer token and TLS) and fixes kthena#1736; another is kthena #1634, making the KV cache GC configurable. The newest are kthena #1812, a design proposal for an OpenAI-compatible Batch API written against a working proof of concept and part of kthena#1291, and kthena #1817, which fixes a double close panic in the session sticky store.

Issues: all issues I authored, including the four part kthena backlog triage in kthena#1532, and issues I responded to.

Reviews: design review on the maxSurge proposal (kthena#1565), the KVCache e2e suite (kthena#1222), and the ModelServing revision history proposal (kthena#1607), plus other PRs I reviewed.

@YaoZengzeng @LiZhenCheng9527, you have reviewed most of my kthena work, would you be willing to sponsor it? Completely fine if the timing is not right, and I will update the checklist either way.

/cc @kube-gopher @FAUST-BENCHOU @hzxuzhonghu