Hacker ApprovedDepartments

etcd

Shares rank 3 of 6 key-value stores by recommenders

Recommended as a distributed key-value store that performs well on fast disks. Critics cite unreliability, data corruption and scaling limits.

Recommended as a distributed key-value store that performs well on fast disks. Critics cite unreliability, data corruption and scaling limits.

32 people gave an opinion

recommend it
9
criticize it
23
approval share
28%
Each figure is one person. RecommendsCriticizes

Recommended in 4 of 13 months

  1. November 2025: recommended
  2. December 2025: recommended
  3. January 2026: not recommended
  4. February 2026: recommended
  5. March 2026: not recommended
  6. April 2026: not recommended
  7. May 2026: not recommended
  8. June 2026: not recommended
  9. July 2026: not recommended
  10. August 2026: recommended
  11. September 2026: not recommended
  12. October 2026: not recommended
  13. November 2026: not recommended (month in progress)

Why it doesn’t carry the badge

  • 9 people recommend itNeeds 10 or more, not met.
  • 0.3 recommenders for every criticNeeds 2 to 1 or better, not met.
  • Recommended in 4 of 13 monthsNeeds 4 or more, met.

The same three rules apply to every product, and the badge can’t be bought. How the badge works

What people say about it

The reasons people give, with the number of comments that give each one. Open a reason to read what they said.

Why people recommend it

Each mark is one comment

  • Handles tens of thousands of QPS

    “It's totally possible to run tens of thousands of QPS on etcd if your disks are NVMEs”

    Comment from 25 Nov 2025 in “Building the largest known Kubernetes cluster”

    “Having run a fleet of 100s of etcd clusters for 10000s of rps”

    Comment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

    “We had run etcd at much larger volumes without issues”

    Comment from 26 Nov 2025 in “Building the largest known Kubernetes cluster”
  • Trusted for distributed key-value storage

    “If you need a full-featured distributed KV store, you should look at etcd.”

    Comment from 21 Nov 2025 in “Show HN:Nkv – a tiny key-value store with pub/sub and keyspaces for shared state”

    “for small projects, SQLite, rqlite, or etcd”

    Comment from 20 Dec 2025 in “Go ahead, self-host Postgres”

    “big projects like Docker, Kubernetes, etcd”

    Comment from 1 Dec 2025 in “You want microservices, but do you need them?”

What people warn about

Each mark is one comment

  • Frequent bugs and unreliability

    “Sounds to me that etcd has data handoff consistency issues, and is just flying by seat of its pants”

    Comment from 21 Feb 2026 in “When etcd crashes, check your disks first”

    “has had numerous liveness and safety bugs, with one happening as recently as December of 2025”

    Comment from 11 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

    “the OG etcd was a disaster that needed better engineers to completely rewrite”

    Comment from 1 Nov 2026 in “Jepsen: ParticleDB 1.1.4”
  • Corruption and data loss reported

    “Etcd corruption and loss of quorum is extremely common in practice”

    Comment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

    “etcd had serious issues including silent data corruption”

    Comment from 11 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”
  • Scaling and size limits

    “if you create too many resources you can easily use up all of the 8GB hard coded maximum size in etcd”

    Comment from 24 Nov 2025 in “Building the largest known Kubernetes cluster”

    “The main issue I've actually encountered is that etcd doesn't scale”

    Comment from 23 Apr 2026 in “I Am Building a Cloud”

Other key-value stores people recommend

The four most recommended other key-value stores, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.

etcd compared with 4 alternatives
ProductRecommendCriticizeApprovalMonthsOwnedBadge
LMDB20969%9—Approved
RocksDB20483%8—Approved
etcdThis page92328%4——
Cloudflare KV91243%8——
LevelDB5550%3——
See all 6 key-value stores, ranked

What people said

37 of the 39 comments that mention it, newest first. Each links to the original comment.

“the OG etcd was a disaster that needed better engineers to completely rewrite”

CriticizesComment from 1 Nov 2026 in “Jepsen: ParticleDB 1.1.4”

“I've always been disappointed with similar systems, etcd including.”

CriticizesComment from 24 Oct 2026 in “Twelve years of distributed SQLite”

“it works quite well for me”

RecommendsComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“etcd is notoriously unreliable and one of the biggest problems in k8s.”

CriticizesComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“its neither performant nor reliable”

CriticizesComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“Their lock, leader election, sessions, and leases are all awful and I'd never recommend anyone to use those.”

MixedComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“Etcd corruption and loss of quorum is extremely common in practice”

CriticizesComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“Having run a fleet of 100s of etcd clusters for 10000s of rps”

RecommendsComment from 12 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“etcd had serious issues including silent data corruption”

CriticizesComment from 11 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“etcd is some of the most amateur code I've ever seen”

CriticizesComment from 11 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“has had numerous liveness and safety bugs, with one happening as recently as December of 2025”

CriticizesComment from 11 Aug 2026 in “Why Go Is an Ideal Language for AI-Assisted Software Engineering”

“I think all big cloud providers replaced etcd with a custom in house implementation in their K8s offerings for this reason”

CriticizesComment from 8 Jul 2026 in “Cloudflare Meerkat - Globally distributed consensus”

“etcd is one of the biggest limitations to Kubernetes in my opinion”

CriticizesComment from 8 Jul 2026 in “Cloudflare Meerkat - Globally distributed consensus”

“spinning up 3 etcd databases”

CriticizesComment from 16 Jun 2026 in “What job interviews taught me about Kubernetes”

“it has a lot of cruft and I'm not a fan”

CriticizesComment from 4 May 2026 in “K3sup – bootstrap K3s over SSH in < 60s”

“Apart from that bunch of issues due to etcd, longhorn.”

CriticizesComment from 4 May 2026 in “K3sup – bootstrap K3s over SSH in < 60s”

“The main issue I've actually encountered is that etcd doesn't scale”

CriticizesComment from 23 Apr 2026 in “I Am Building a Cloud”

“Sounds to me that etcd has data handoff consistency issues, and is just flying by seat of its pants”

CriticizesComment from 21 Feb 2026 in “When etcd crashes, check your disks first”

“we should have found the proper parameters to make etcd wait for quorum”

CriticizesComment from 21 Feb 2026 in “When etcd crashes, check your disks first”

“That’s a design issue in etcd.”

CriticizesComment from 21 Feb 2026 in “When etcd crashes, check your disks first”

“A lot of k8s setups would be better served by just hitting postgres”

MixedComment from 21 Feb 2026 in “When etcd crashes, check your disks first”

“the outcome described in the article is pretty nasty”

CriticizesComment from 21 Feb 2026 in “When etcd crashes, check your disks first”

“etcd and k8s just seem to be straight up mismanaged”

CriticizesComment from 1 Feb 2026 in “"Giving up upstream-ing my patches & feel free to pick them up"”

“It is impossible to get someone from the project to review my PRs”

CriticizesComment from 31 Jan 2026 in “"Giving up upstream-ing my patches & feel free to pick them up"”

“rather than running/securing/maintaining just another etcd cluster”

CriticizesComment from 21 Dec 2025 in “Go ahead, self-host Postgres”

“for small projects, SQLite, rqlite, or etcd”

RecommendsComment from 20 Dec 2025 in “Go ahead, self-host Postgres”

“Etcd is not a requirement.”

CriticizesComment from 8 Dec 2025 in “Proxmox Datacenter Manager 1.0 available”

“etcd is definitly not designed for use across large distances.”

CriticizesComment from 4 Dec 2025 in “Proxmox Datacenter Manager 1.0 available”

“big projects like Docker, Kubernetes, etcd”

MixedComment from 1 Dec 2025 in “You want microservices, but do you need them?”

“We had run etcd at much larger volumes without issues”

RecommendsComment from 26 Nov 2025 in “Building the largest known Kubernetes cluster”

“It's definitely not a recovery strategy for a full etcd by itself”

CriticizesComment from 26 Nov 2025 in “Building the largest known Kubernetes cluster”

“It's totally possible to run tens of thousands of QPS on etcd if your disks are NVMEs”

RecommendsComment from 25 Nov 2025 in “Building the largest known Kubernetes cluster”

“it wasn't built for scale or for speed.”

CriticizesComment from 24 Nov 2025 in “Building the largest known Kubernetes cluster”

“if you create too many resources you can easily use up all of the 8GB hard coded maximum size in etcd”

CriticizesComment from 24 Nov 2025 in “Building the largest known Kubernetes cluster”

“Etcd is the main bottleneck in upstream k8s”

CriticizesComment from 24 Nov 2025 in “Building the largest known Kubernetes cluster”

“is going to run circles around etcd for kube controlplane operations”

CriticizesComment from 24 Nov 2025 in “Building the largest known Kubernetes cluster”

“If you need a full-featured distributed KV store, you should look at etcd.”

RecommendsComment from 21 Nov 2025 in “Show HN:Nkv – a tiny key-value store with pub/sub and keyspaces for shared state”

Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.