All Versions
19
Latest Version
Avg Release Cycle
27 days
Latest Release
53 days ago

Changelog History
Page 1

  • v271.0.0

    April 09, 2020

    ➕ Additions

    • 🐎 Merged a PR that changes the NATS gem from the EventMachine-based to 'pure ruby'. For most operators, this will have no impact. For operators of large foundations, performance and TLS-related errors could be seen with the EM-based gem which the pure-ruby gem relieves. Thanks to @beyhan@friegger @langered @videlov#2254

    🛠 Fixes

    • ⏱ thanks for fixing an issue with the scheduled tasks clean up @beyhan#2256
    • thanks for adding the missing option @beyhan #2248

    🔒 Security

  • v270.12.0

    February 26, 2020

    🆕 New Features

    • 👌 Support index or id for ignore command (#170702879)
    • ➕ Add support for running metrics server in local mode (#2235)

    🐛 Bug Fixes

    • ✂ Delete tasks on a schedule instead of "as-queued" (#170964112)
    • 👍 Metrics server supports cloud configs with no networks definition (#171253546)
    • 🛠 Fix warning about running errands on multiple instances (#170114966)
    • Only copy blobstore config file to VMs once (#169432981)
  • v270.11.0

    January 16, 2020

    🔋 Features

    • 👍 Allow operators to turn on host style URLs for s3 blobstores (#170090112)
    • ⚡️ Update Director response so that bosh vms shows stemcell version (#169015295)
    • 🔦 Expose dry-run functionality for clean-up (#167711536)
    • ➕ Add keep-orphaned-disks for clean-up command (#170309487)
    • ⚡️ Metrics Updates (##4428257)
      • Expose number of IPs and free IPs per network
      • Expose number of unresponsive agents
      • Expose number and state of tasks
      • Expose global resurrector status
    • 👌 Improve blobstore performance by changing default type blobstore nginx type to application/octet-stream (#170268644)
      (#170122072)

    🐛 Bug Fixes

    • Stop removing network information from errand VMs (#166359647)
  • v270.10.0

    December 06, 2019

    🔋 Features

    • Include VM CID in agent ping error message (#169682585)
    • ➕ Add support for customer-provided encryption keys with signed URLs on GCS (#168148185)
    • ➕ Add prometheus metrics for director api (#169331435)
      • overall metrics work still in progress
    • 0️⃣ Default blobstore nginx workers to auto (#169988440)
    • 🚀 Update VM/disk metadata when deploy tags change without need for --recreate (#168917397)

    🐛 Bug Fixes

    • ⚡️ Skip full instance update iff DNS or tags changed (#168917397)
    • 🛠 Fix nginx.conf not using SSL for ipv6 listen (#169694744)
  • v270.9.0

    November 15, 2019

    🔋 Features

    • When enabled, metrics endpoint is protected by mTLS and publicly accessible (#169331194)

    🐛 Bug Fixes

    • Remove overly harsh credential validation in s3cli when using env_or_profile(#2219)
    • 🚀 Don't fail clean-up if a runtime config does not contain releases (#169220410)
  • v270.8.0

    October 31, 2019

    🔋 Features

    • 👌 Improve performance of /deployments endpoint (#168545859)
    • 🚀 Introduce support for signed URLs on BOSH deployed VMs for accessing the blobstore
      • Currently supports internal webdav blobstore and S3 without user provided encryption keys
      • Upcoming releases will support GCS blobstore and user provided keys
    • Introduce /metrics endpoint, enabled via director.metrics_server.enabled property
      • Currently this is only exposed locally on the director
      • More metrics and functionality will be added in upcoming releases
      • Add new /resurrection endpoint to return status of global resurrection (#168964490)
      • Increase timeout on release lock (#168883842)
    • Re-add top level properties

    🐛 Bug Fixes

    • 🛠 Fix issue where sync-dns exits unexpectedly yielding errand VMs to not have have a records.json file (#169302667)
    • ✂ Remove duplicate ruby package
  • v270.7.0

    October 04, 2019

    🔋 Features

    • Include tags on vm creation. This change also includes tags in the bosh env hash for CPIs that
      wish to tag VMs on creation. (#167657971)

    🐛 Bug Fixes

    • 👷 blobstore.secret property is not required in the director job (#168853977)
  • v270.6.0

    September 30, 2019

    🔋 Features

    • ⚠ Show warning when running errands across multiple instances (#166485688)
    • 🚀 Continue v1 manifest removal: Remove properties at deployment and instance_group level (#167907044)
    • 🚀 Enable parallel first time deploy across AZs via initial_deploy_az_update_strategy (#168067218)
    • ➕ Add ability to generate signed URLs for blobstore clients – not yet used (#167978684)

    🐛 Bug Fixes

    • 🚀 Don't emit scan and fix tasks for a deployment while it is being deployed (#167502252)
    • Director API now returns static & dynamic IPs associated with instances (#2206)
    • 🚀 Do not consider obsolete instance plans when selecting reusable VMs during deploy (#168414135)
  • v270.5.0

    August 12, 2019

    🔋 Features

    • 🚀 bosh clean-up --all does not remove releases referenced in a runtime config (#167737389)
    • ⚡️ Update Last Activity At to Finished At to more accurately represent task timestamps (#167549610)
    • More improvements to new start, stop, recreate, restart commands

    🐛 Bug Fixes

    • 💎 Show AZ name in task debug log rather than Ruby object (#166847303)
    • 👷 Properly shut down worker children when stopping Director workers (#163419348)
  • v270.4.0

    July 22, 2019

    🔋 Features

    • ➕ Add new endpoints for restarting and recreating an instance without impacting
      🚀 the entire deployment (#167149815) (#167149818)
    • Tasks associated with new start, stop, recreate, and restart endpoints show up in tasks table
      0️⃣ by default (#167182280)
    • ➕ Add update_mode as a key for each variable (#164818604)

    🐛 Bug Fixes

    • 🛠 Fix edge case of director giving away IP addresses (#167185510)
    • 🛠 Fix CCK should not only update links resolution when necessary (#167268961)??????