Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Browse files
Browse the repository at this point in the history
107297: storage,kvserver: Foundational changes for disaggregated ingestions r=sumeerbhola a=itsbilal This change contains two commits (split off from the original mega-PR, #105839). The first is a pkg/storage change to add new interface methods to call pebble's db.ScanInternal as well as implement related helper methods in sstable writers/batch readers/writers to be able to do disaggregated snapshot ingestion. The second is a kvserver/rditer change to allow finer-grained control on what replicated spans we iterate on, as well as to be able to specifically opt into skip-shared iteration over the user key span through the use of `ScanInternal`. --- **storage: Update Engine/Reader/Writer interfaces for ScanInternal** This change updates pkg/storage interfaces and implementations to allow the use of ScanInternal in skip-shared iteration mode as well as writing/reading of internal point keys, range dels and range keys. Replication / snapshot code will soon rely on these changes to be able to replicate internal keys in higher levels plus metadata of shared sstables in lower levels, as opposed to just observed user keys. Part of #103028 Epic: none Release note: None **kvserver: Add ability to filter replicated spans in Select/Iterate** This change adds the ability to select for just the replicated span in rditer.Select and rditer.IterateReplicaKeySpans. Also adds a new rditer.IterateReplicaKeySpansShared that does a ScanInternal on just the user key span, to be able to collect metadata of shared sstables as well as any internal keys above them. We only use skip-shared iteration for the replicated user key span of a range, and in practice, only if it's a non-system range. Part of #103028. Epic: none Release note: None 108336: sql: retry more distributed errors as local r=yuzefovich a=yuzefovich This PR contains a couple of commits that increase the allow-list of errors that are retried locally. In particular, it allows us to hide some issues we have around using DistSQL and shutting down SQL pods. Fixes: #106537. Fixes: #108152. Fixes: #108271. Release note: None 108406: server,testutils: remove complexity r=yuzefovich,herkolategan a=knz There is a saying (paraphrasing) that it always takes more work removing unwanted complexity than it takes to add it. This is an example of that. Prior to this commit, there was an "interesting" propagation of the flag that decides whether or not to define a test tenant for test servers and clusters. In a nutshell, we had: - an "input" flag in `base.TestServerArgs`, which remained mostly immutable - a boolean decided once by `ShouldStartDefaultTestTenant()` either in: - `serverutils.StartServerOnlyE` - or `testcluster.Start` - that boolean choice was then propagated to `server.testServer` via _another_ boolean config flag in `server.BaseConfig` - both the 2nd boolean and the original input flag were then again checked when the time came to do the work (in `maybeStartDefaultTestTenant`). Additional complexity was then incurred by the need of `TestCluster` to make the determination just once (and not once per server). This commit cuts through all the layers of complexity by simply propagating the choice of `ShouldStartDefaultTestTenant()` back into the `TestServerArgs` and only ever reading from that subsequently. Release note: None Epic: CRDB-18499 108465: cloudccl: allow external connection tests to be run in parallel r=rhu713 a=rhu713 Currently external connection tests read and write to the same path in cloud storage. Add a random uint64 as part of the path so that test runs have unique paths and can be run in parallel. Fixes: #107407 Release note: None 108481: acceptance: stabilize start-single-node in tcl test r=santamaura a=dhartunian We've continued to see flakes on this test which contain messages of throttled stores on node startup. The hypothesis is that these are due to leftover data directories from prior startups during the same test. This change clears the `logs/db` data directory for those invocations and also adds the sql memory flag which the common tcl function also uses. Resolves #108405 Epic: None Release note: None 108496: kv: unit test `PrepareTransactionForRetry` and `TransactionRefreshTimestamp` r=miraradeva a=nvanbenschoten Informs #104233. This commit adds a pair of new unit tests to verify the behavior of `PrepareTransactionForRetry` and `TransactionRefreshTimestamp`. These functions will be getting more complex for #104233, so it will be helpful to have these tests in place. The tests also serve as good documentation. Release note: None Co-authored-by: Bilal Akhtar <bilal@cockroachlabs.com> Co-authored-by: Yahor Yuzefovich <yahor@cockroachlabs.com> Co-authored-by: Raphael 'kena' Poss <knz@thaumogen.net> Co-authored-by: Rui Hu <rui@cockroachlabs.com> Co-authored-by: David Hartunian <davidh@cockroachlabs.com> Co-authored-by: Nathan VanBenschoten <nvanbenschoten@gmail.com>
- Loading branch information
