Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Patch release 3.2.5 #3653

Merged
merged 10 commits into from
Dec 6, 2023
Merged

Patch release 3.2.5 #3653

merged 10 commits into from
Dec 6, 2023

Commits on Dec 6, 2023

  1. docs: update s3 compatible storage description with information about… 

    … custom CAs
    @karol-kokoszka
    karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    10dcead View commit details
    Browse the repository at this point in the history

  2. add(util): introduced worker pool

    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    c8ecb64 View commit details
    Browse the repository at this point in the history

  3. add(repair): use worker pool 

    This change makes it easier to introduce changes to repair parallelism,
like bypassing 1 job per 1 node rule.
    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    ad712eb View commit details
    Browse the repository at this point in the history

  4. dockerfile: add apt-get upgrade during build 

    Similar to scylladb/scylladb#16222 ,we need to
make sure our docker image is updated to latest for all OS packages

Fixes: #3646
    @yaronkaikov @karol-kokoszka
    yaronkaikov authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    1aa18d0 View commit details
    Browse the repository at this point in the history

  5. fix(codeowners): dist part for release engineering

    @karol-kokoszka
    karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    2484e54 View commit details
    Browse the repository at this point in the history

  6. go.mod: use rclone which accumulates old transfer statistics 

    Currently, SM disables transfer pruning (MaxCompletedTransfers = -1) because it needs complete upload statistics. This is the main reason for memory leak detected in issue #3298. With this change, transfers can be purged while preserving total upload statistics exposed at the core/aggregated endpoint.
    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    5f46177 View commit details
    Browse the repository at this point in the history

  7. fix(rc): use aggregated stats for job progress 

    Recent changes in used rclone fork changed upload statistics aggregation. Total upload statistics are now available at core/aggregated endpoint.
    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    2328434 View commit details
    Browse the repository at this point in the history

  8. add(scylla_client_test): test RcloneJobProgress uploaded/skipped bytes

    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    06c4612 View commit details
    Browse the repository at this point in the history

  9. fix(rclone): don't use MaxBacklog 

    Uploading much more files than the MaxBacklog forces rclone to keep references to uploaded objects even when there is no need for it. This partially results in memory leak detected in #3298.
    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    72c0670 View commit details
    Browse the repository at this point in the history

  10. fix(repair): schedule and run repair on single node cluster 

    SM 3.2 changed repair param validation so that scheduling and running repair of a single node cluster is treated as an error. Although this is acceptable in terms of real users, the ability to schedule/run repair on a simple single node cluster is useful for tests and some experiments.

Fixes #3649
    @Michal-Leszczynski @karol-kokoszka
    Michal-Leszczynski authored and karol-kokoszka committed Dec 6, 2023
    Configuration menu
    Copy the full SHA
    1b3e746 View commit details
    Browse the repository at this point in the history