Cases9 cases

4 to 8 of 9

1. IL-0020 During an EKS 1.30 upgrade the descheduler tried to reschedule every pod on the cordoned nodes at once and production workloads went down

   UK Ministry of Justice2025

   Symptoms

   At 09:36 on 18 March 2025, during the upgrade of the live cluster to EKS 1.30, users reported that their pods were being terminated and were not restarting automatically, in both production and non-production environments. Automated infrastructure alerts fired as well, and production services had outages.

   Pod evictionPods PendingService outageUpgrade stall

   Trigger

   Upgrading the live EKS cluster from 1.29 to 1.30, during which the descheduler attempted to reschedule all pods on the cordoned nodes simultaneously.

   Root cause

   Faulty deployment

   Component

   DeschedulerNode poolCluster autoscaler

   Contributing factor

   No headroomNo alert on limitUnreviewed defaults

   Detection

   Users reported at 09:36 that pods were being terminated and not restarting, and automated infrastructure alerts fired.

   Remediation

   • Removed the cordon annotation so pods could be rescheduled
   • Deleted stuck 1.30 nodes to force their recreation
   • Added a new node group of 20 nodes running 1.30 by hand
   • Cordoned and drained all old 1.29 nodes onto the new 1.30 nodes

   Action items

   • Raise an AWS Production System Down support case (Case ID: 174229643100496)
   • Get guidance from AWS support on querying control-plane logs to find the root cause of similar issues
   • Document a runbook for emergency scaling of a new node group in case of unexpected capacity shortages
   • Alert on an imbalance between the autoscaling group and the Kubernetes nodes
   • Review the descheduler configuration to understand why it throws pods off cordoned nodes, and fix it where possible
   • Alert on spikes in the pending pod count

   Lessons

   The write-up lists review actions rather than lessons. They show the team had no documented way to add an emergency node group, no alert on a mismatch between the autoscaling group and the nodes Kubernetes sees, no alert on a spike in pending pods, and no clear understanding of how the descheduler treats cordoned nodes.

   View case
2. IL-0084 Packet drops on a bare metal provider's vSwitch forced Namespace to rebuild its fra3 region and left global capacity short, and the next day slow NVMe writes on a control plane node drove the Kubernetes API server to very high CPU

   Namespace2024

   Symptoms

   From 17 October 2024, packets through the vSwitch that joined Namespace's bare metal machines at one provider in the fra3 region were dropped at random. Retries in TCP, QUIC and the Namespace control plane hid most of it, but the SLO error budget for latency targets on key interactions began to burn.

   Intermittent connection failuresHigh latencyControl plane lagService outage

   Trigger

   Random packet drops on the bare metal provider's vSwitch from 17 October 2024; for the second outage, a rate of Kubernetes changes that the slow NVMe on one control plane node could no longer keep up with.

   Root cause

   Cloud provider outage

   Component

   Kubernetes API server

   Contributing factor

   No capacity planningNo headroomOverlapping failures

   Detection

   SLO-based monitoring of latency targets showed the error budget burning while retries hid the individual packet drops.

   Remediation

   • diverted all customer traffic from fra3 to other regions
   • connected each provider machine to the zrh1 cluster over Wireguard instead of the vSwitch
   • repaved every fra3 machine with a new identity, certificates and encryption-at-rest keys, dropping regional caches to recover faster
   • cloned existing datasets in a more distributed way to speed up repaving the machines
   • added machines back to the new region in batches and returned the region to the compute pool
   • redirected workloads away from zrh1 during the second outage
   • cloned the failing control plane node's key state to a higher-capacity node that took over its role and IP address

   Action items

   • Credit all instances that ran during the outages back to customers
   • Build up backup regions and solutions so a similar event has options that do not affect customers

   Lessons

   Running hot with a delayed hardware delivery left no room to take a region out, so fixing a provider network fault turned into a global capacity shortage. After one outage the team was anchored on its cause and first looked for network problems in the second outage, which had a different cause: a control plane node whose disk could not keep up with the Kubernetes change rate.

   View case
3. IL-0024 Expensive cache-backed API calls and uncancelled database requests overloaded MongoDB and took the Hugging Face Hub down three times

   Hugging Face2024

   Symptoms

   At 08:45 CET on 22 April 2024 the Hugging Face Hub became unstable, and by 09:30 it was unresponsive. For most users the website was slow or could not be reached at all.

   Service outageHigh latencyDatabase overload

   Trigger

   Bursts of requests to revision endpoints that read heavily from the *_info_cache collections: a call to https://huggingface.co/api/datasets/lighteval/mmlu/revision/main rose sharply right before the first crash, and a call to https://huggingface.co/api/models/sentence-transformers/all-mpnet-base-v2/revision/main, a model with 300+ associated spaces, lined up with the crash at 21:30 on 23 April.

   Root cause

   Software defect

   Component

   Application databaseModel hubCluster autoscaler

   Contributing factor

   No headroomRetry stormOverlapping failures

   Detection

   Instability started at 08:45 on 22 April and the Hub was unresponsive by 09:30; the team followed MongoDB Atlas metrics and a MongoDB support case. The root cause was found days later from a new graph of requests per URL.

   Remediation

   • Added a firewall rule blocking all requests with a Java/* user agent
   • Disabled the quicksearch feature on the Hub
   • Scaled the MongoDB Atlas cluster by hand from tier 2 to tier 3
   • Switched the Hub to maintenance mode with a 503 page through AWS WAF and blocked incoming traffic
   • Scaled all Hub pods down to 0 and brought them back
   • Set an aggressive rate limit
   • Added timeouts to requests and reworked database queries that were slow or returned large bodies

   Action items

   • Add timeouts to requests based on their usual completion time
   • Optimise requests that took too long or returned a large body, and rework poorly designed database queries
   • Remove old unused indexes from the database
   • Build new dashboards and monitoring metrics
   • Move the *_info_cache collections to Redis
   • Make it possible to abort any database request when the client aborts theirs

   Lessons

   Finding the root cause took two to three days: the existing charts did not show it, and a per-URL request breakdown did. Not cancelling database work when a client goes away turns a strained database into a thundering herd. A close inspection of the database exposed architectural flaws and bottlenecks, such as a collection with large unused indexes, and taught the team which metrics need closer watching to anticipate crashes. Unrelated failures happening at the same time added to the stress and made restoring service harder.

   View case
4. IL-0123 A response buffer allocated anew in a loop made OpenAI's routing layer nodes hit their memory limits and fail readiness checks on a morning with more completions than any day before, causing 502 and 503 errors across all models

   OpenAI2023

   Symptoms

   Between 5:42 AM and 7:16 AM PT on 8 November 2023 a large share of requests to OpenAI failed with 502 or 503 errors. Every model and API endpoint saw significant failures.

   High error rateService outageProbe failures

   Trigger

   A morning with dramatically more completions than any previous day pushed the routing layer nodes, which already had a chronic memory problem, past their memory limits.

   Root cause

   Software defect

   Component

   Model request routingContainer memory limit

   Contributing factor

   No alert on limitNo headroom

   Remediation

   • limited incoming traffic
   • redeployed the routing service en masse
   • ramped traffic back up slowly

   Action items

   • Pre-allocate the responseBuffer and reuse it instead of allocating a new one in a loop (3X improvement in memory and CPU usage)
   • Adjust the configured memory limits to an appropriate level so the service has significant headroom
   • Add rate limit controls to shed traffic more gracefully
   • Increase the service's capacity
   • Change alerting to catch the underlying memory behaviour before it escalates into service issues
   • Configure autoscaling for the service, which was not possible before because added capacity could hurt an upstream service

   View case
5. IL-0115 One of OpenAI's three production AKS clusters went offline through a human error, and rebuilding it took 10½ hours because an AKS Kubernetes upgrade broke the ingress and AKS cannot be downgraded

   OpenAI2022

   Symptoms

   At 18:56 PDT on 23 June 2022 (01:56 UTC on 24 June) several models went offline at once: some fine-tuned curie and davinci models, the Codex beta models code-davinci-001 and code-cushman-001, the legacy GPT models curie, babbage and ada, and all Embeddings models.

   Service outageHigh error rateIntermittent timeouts

   Trigger

   A destructive operation on one of the three production clusters, caused by human error that the existing safeguards did not catch, took the whole cluster offline.

   Root cause

   Operator error

   Component

   Ingress controllerModel hub

   Contributing factor

   No safeguard on destructive changeNo headroomDeferred upgradesOverlapping failures

   Remediation

   • re-allocated the models from the lost cluster to the remaining two clusters
   • rebuilt the VPN link with Azure support after losing the connection to AKS during cluster setup
   • built workarounds with Azure support after losing the connection to Azure Private Link
   • tore down the cluster and rebuilt it from scratch after the upgraded Kubernetes version broke the ingress
   • moved traffic back into the rebuilt cluster between 05:30 and 06:30 PDT

   Action items

   • Add checks and locks so that this and similar production-critical infrastructure cannot be modified by accident
   • Add better visibility into the operations being performed and make parts of the system more lexicographically distinct
   • Keep working with Azure on the AKS network connectivity issues so they do not disrupt new cluster operations
   • Update playbooks with the new cluster failure modes and change the setup steps for new clusters and infrastructure
   • Upgrade Kubernetes systematically and set aside time so the clusters do not fall behind again
   • Shard fine-tuned models across multiple clusters and work towards more automatic cross-cluster replication
   • Distribute and replicate Azure Blob Storage so fine-tuned models load faster and more reliably
   • Finish rolling out node-to-node caching to relieve Azure Blob Storage bottlenecks
   • Change how model weights are stored so they are more efficient and faster to load

   Lessons

   Losing one of three clusters exhausted the GPU headroom, and GPUs could not be added on short notice, so full recovery depended on bringing the same cluster back. Rebuilding it exposed a deferred Kubernetes upgrade: the newer version broke the ingress, and because AKS cannot downgrade there was no way back except a full rebuild. Recovering many large models at once is limited by the bandwidth of the storage they load from, and the caches that normally hide this do not help when everything moves to a different cluster.

   View case

By root cause7