76611: roachpb: add lock table metadata structures r=AlexTalks a=AlexTalks This change adds the protobuf structures, as well as the method, needed to capture the state of a replica's lock table. This is part of the work coming out of #75541, and is needed to be able to implement the `QueryLocks` RPC. Release justification: (Category 2) The structures created here are entirely new, for the purposes of observability, and do not represent a risk to any existing code or functionality. Release note: None 76897: release: use bazel to build release binaries r=rail,jlinder a=rickystewart This is a pretty large swath of related changes that update the release pipeline to use Bazel for builds. 1. Update `.bazelrc` to add a new `*base` config for each cross-config. The existing cross configs will continue to work as before. Each `*base` config has the same contents as the corresponding cross config EXCEPT the `--workspace_status_command` argument is missing. 2. Stamping. `build/bazelutil/stamp.sh` now takes up to 3 more arguments in addition to the target triple: `build-channel`, `build-tag`, and `build-type`. These are documented in the script and have appropriate defaults for development builds if not specified. 3. Refactor `pkg/release`; the new API is more high-level rather than allowing injecting arbitrary `gotags`, etc. In the future it won't be impossible to add this functionality back in if we want it. 4. Prepare to fork the `Make and Publish Build` build configuration in TC by moving its script into `build/teamcity` and update the script to use Bazel. 5. Update `bazci` to make sure it reads the Windows `geos` libraries from the appropriate location in `bazel-bin`. 6. Update `publish-artifacts` and `publish-provisional-artifacts` to use the new `pkg/release` API. Closes #67147. Closes #67161. Closes #67279. Closes #67322. Release note (general change): Use Bazel to build release binaries Release justification: update release process 77247: server: Allow bypass of tsdump's strict checks r=abarganier a=rimadeodhar tsdump performs strict checks that verify that it has data for exactly the stores it expects to see according to the node->store mapping. It can happen that there legitimately isn't data there, for example when nodes are down during the tsdump window. We still want the strict checks, but now tsdump will defer returning an error and allow restarting with a "-" in lieu of the tsimport file name to display the data anyway. Additionally, cluster settings are set via SET CLUSTER SETTING as opposed to overriding settings directly. Release note (cli change): debug tsdump command allows viewing timeseries data even in case of node failures by rerunning the command with the import filename set to "-". Addresses #75993 Release justification: low risk, high benefit changes to existing functionality 77483: sql/catalog: add return error to RunPostDeserializationChanges r=ajwerner a=rafiss Release justification: low risk enhancement to existing functionality. Release note: None 77493: sql: make session_revival_token.enabled tenant-ro r=otan,jaylim-crl a=rafiss I was hoping to wait for the new cluster setting syntax to be completed, but since it's getting close to the branch cut time I'd rather merge this now so we don't forget at the last minute. Release justification: low risk change to new functionality. Release note: None Co-authored-by: Alex Sarkesian <sarkesian@cockroachlabs.com> Co-authored-by: Ricky Stewart <ricky@cockroachlabs.com> Co-authored-by: rimadeodhar <rima@cockroachlabs.com> Co-authored-by: Rafi Shamim <rafi@cockroachlabs.com>

ci: duplicate Internal / Release / Process / Make and Publish Build CI job for Bazel #67161

Description

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

ci: duplicate `Internal / Release / Process / Make and Publish Build` CI job for Bazel #67161