0.2.1

• Bump github.com/aws/aws-sdk-go from 1.43.10 to 1.43.11 (#131)
• Bump k8s.io/client-go from 0.19.3 to 0.23.4 (#119)
• Bump golang to 1.17.8, security fix for issues/51112

image:
Docker Hub: zmazay/s3sync-service:0.2.1
Quay.io: quay.io/s3sync-service/s3sync-service:0.2.1

0.2.0

• support S3 compatible services
• upgrade golang to 1.17
• upgrade base alpine image to 3.15.0
• support Apple silicon

image:
Docker Hub: zmazay/s3sync-service:0.2.0
Quay.io: quay.io/s3sync-service/s3sync-service:0.2.0

0.1.2

• Symlinks within the sync directory are skipped from processing (#100)
• Introduced forced reload feature, which might be useful for cleaning/retrying to sync all the failed files without the application restart. The forced reload can be triggered with an HTTP request, eg. - curl http://127.0.0.1:8090/reload?force. For more details please check - https://docs.s3sync-service.org/en/0.1.2/how-it-works/#application-reload
• Upgraded to Go 1.16
• Improved the initial sync by switching to a more efficient WalkDir method
• Numerous internal improvements, tests, etc.

image:
Docker Hub: zmazay/s3sync-service:0.1.2
Quay.io: quay.io/s3sync-service/s3sync-service:0.1.2

0.1.1

• All the site metrics are reset to 0 on reload
• Introduce quay.io mirror

Breaking changes:

• The s3sync_errors_count metrics type is changed from Counter to Gauge

image:
Docker Hub: zmazay/s3sync-service:0.1.1
Quay.io: quay.io/s3sync-service/s3sync-service:0.1.1

0.1.0

• Fixed base image for multi-arch images, previously sh or any other command (apart from s3sync-service) was failing on anything, not linux/amd64
• Implemented application reload - no need to restart if the config has changed, check docs for more details
• Implemented graceful shutdown - workers will try to finish currently running task before the shutdown
• (#67) Introduced Helm Charts, check the guidelines on the usage
• (#60) k8s configmap can be used as a source of the configuration, the application will be reloaded if configuration changes detected
• Introduced simple HTTP server for triggering reload and fetching some basic info - https://docs.s3sync-service.org/en/0.1.0/http-server/
• Fixed #65 - S3 object size wasn't fetched properly under some circumstances

Breaking changes

• The watch_interval now expects time units for the value, valid time units are ns, us (or µs), ms, s, m, h.

image: zmazay/s3sync-service:0.1.0

0.0.14

• Upgraded to golang 1.15.2
• Added darwin binary build to the release flow
• Upgraded golang dependencies:
  • aws-sdk-go v1.26.8 -> v1.35.5
  • prometheus/client_golang v1.1.0 -> v1.7.1
  • prometheus/client_model v0.1.0 -> v0.2.0
  • prometheus/common v0.7.0 -> v0.14.0
  • prometheus/procfs v0.0.8 -> v0.2.0
  • logrus v1.4.2 -> v1.7.0
  • yaml.v2 v2.2.7 -> v2.3.0

image: zmazay/s3sync-service:0.0.14

0.0.13

• Introduced s3_ops_retries parameter to have a certain amount of retries before throwing an error
• Logging improvements:
  • added app and version fields
  • fields rename for logstash/filebeat compatibility:
    • ts -> @timestamp
    • msg -> message
• Check if run in k8s if true set checksum_workers: 2 otherwise checksum_workers: CPUCORES * 2

image: zmazay/s3sync-service:0.0.13

0.0.12

• Fixed broken multi-arch images #52

image: zmazay/s3sync-service:0.0.12

0.0.11

• Fixed bug reported in #29
• Upgrade to golang 1.14.1
• Docker images are now multi-arch with the following options - linux/amd64, linux/arm/v6, linux/arm/v7, linux/arm64/v8, linux/386, linux/ppc64le

image: zmazay/s3sync-service:0.0.11

0.0.11a

• Fix #29
• Close #30
• Upgrade to golang 1.14.1

image: zmazay/s3sync-service:0.0.11a