Who contributed to HashiCorp on May 10, 2026?

1 developer shipped this update, including hashi-divyansh.

What were the notable HashiCorp updates?

Fix winner selection, checks outside schedule window should not participate in the winner selection process, Fix winner selection, checks outside schedule window should not participate in the winner selection process (#1270), and nightly.

@hashicorp

Terraform, Vault, Consul - infra-as-code for ops teams

github ↗

Pick a date

Topics: Go Infrastructure Full archive →

The Wire · Showcase

NOMAD AUTOSCALER FIXES CRITICAL WINNER SELECTION BUG

By RepoJournal · Filed 06:01 UTC on May 10, 2026 · About HashiCorp

1 person shipped this

hashi-divyansh @hashi-divyansh 1 cited

Nomad Autoscaler now properly excludes out-of-window checks from scaling decisions, preventing phantom participants from skewing autoscaling logic.

The fix [1] addresses a logic error where checks scheduled outside their window or marked for cancellation were still influencing which checks won the selection process. This could cause the autoscaler to make scaling decisions based on incomplete or stale data. The solution introduces explicit sentinel errors—specifically `errCheckOutsideSchedule`—that mark non-participating checks and filter them from winner selection calculations [2]. The team also refactored schedule validation into the SDK layer and added regression tests to prevent this from slipping back in. Nightly builds across Nomad Autoscaler [3], the Podman driver [4], and Nomad Pack [5] are live for early testing.

Action items

→ Review winner selection logic in your custom scaling policies hashicorp/nomad-autoscaler [plan]
→ Test nightly builds against your Nomad cluster to catch regressions early hashicorp/nomad-autoscaler [monitor]

References

[1] Fix winner selection, checks outside schedule window should not participate in the winner selection process ↗ hashicorp/nomad-autoscaler
[2] Fix winner selection, checks outside schedule window should not participate in the winner selection process (#1270) hashicorp/nomad-autoscaler
[3] nightly ↗ hashicorp/nomad-autoscaler
[4] nightly ↗ hashicorp/nomad-driver-podman
[5] nightly ↗ hashicorp/nomad-pack

Quick answers

What shipped in HashiCorp on May 10, 2026?: Nomad Autoscaler now properly excludes out-of-window checks from scaling decisions, preventing phantom participants from skewing autoscaling logic. In total, 1 commits, 1 pull requests, and 3 releases landed.
Who contributed to HashiCorp on May 10, 2026?: 1 developer shipped this update, including hashi-divyansh.
What were the notable HashiCorp updates?: Fix winner selection, checks outside schedule window should not participate in the winner selection process, Fix winner selection, checks outside schedule window should not participate in the winner selection process (#1270), and nightly.

TERRAFORM PROVIDERS EXPAND GOOGLE CLOUD COVERAGE AND IN-PLACE UPDATES

HashiCorp shipped in-place VCS connection updates for registry modules while Google provider gains traffic config for Vertex AI and promotes NCC gateway resources to GA.

go 19 shipped 1-min read

@hashicorp 1 day ago

VAULT LDAP GETS ON-DEMAND PASSWORD ROTATION, PACKER CUTS v1.16.0

Terraform provider for Vault now supports rotate-on-read for LDAP secrets, letting teams rotate credentials at read time instead of waiting for scheduled intervals.

go 8 shipped 1-min read

@hashicorp 2 days ago

CONSUL FIXES FLOODED MEMBER REJOIN BUG AS VSO PATCHES GRPC FRAME FLOOD

Consul patches a critical WAN flood regression that was incorrectly treating failed members as alive during rejoin sequences.

go 8 shipped 1-min read

@hashicorp 3 days ago

TERRAFORM AWS PROVIDER GAINS S3 TABLES AND OPENSEARCH FEDERATION SUPPORT

The AWS provider shipped three enhancement PRs overnight that expand infrastructure-as-code coverage for S3 Tables metadata properties and OpenSearch security configurations, while the docs team cleaned up deprecated tutorial collections and fixed formatting bugs across Vault and PKI documentation.

go 60 shipped 1-min read

Elsewhere on the wire

AI Agents about 9 hours ago

CLAUDE OPUS 5 LANDS ACROSS THE STACK

The newest Anthropic model is now live in langchain, Cline, and llama-index, with native support for extended reasoning and 1M context windows.

ai-agents 28 shipped 1-min read

Local LLMs about 9 hours ago

OLLAMA LANDS LAGUNA SUPPORT AND CRUSHES MEMORY LEAKS WHILE SGLANG HITS V0.5.16 WITH CONFIDENCE-DRIVEN SPECULATIVE DECODING

Ollama shipped three critical performance and reliability fixes for Metal residency and concurrent access patterns, while SGL-Lang released 0.5.16 with a new speculative algorithm hitting 383.7 tok/s on DeepSeek-V4.