package dolt.services.remotesapi.v1alpha1

Get desktop application:
View/edit binary Protocol Buffers messages

• rpc GetRepoMetadata (GetRepoMetadataRequest, GetRepoMetadataResponse)

  chunkstore.proto:24

  message GetRepoMetadataRequest

  chunkstore.proto:204

  • optional RepoId repo_id = 1

  • optional ClientRepoFormat client_repo_format = 14

  • string repo_token = 2

  • string repo_path = 3

  message GetRepoMetadataResponse

  chunkstore.proto:235

  • string nbf_version = 1

    Version string of the noms binary format for this repository. See types.NomsBinFormat.

  • string nbs_version = 2

    Version string of the nbs format of this repository. See nbs.StorageVersion.

  • uint64 storage_size = 3

    Approximate number of bytes required for storage of all currently-referenced repository table files.

  • string repo_token = 4

  • PushConcurrencyControl push_concurrency_control = 5

• rpc HasChunks (HasChunksRequest, HasChunksResponse)

  chunkstore.proto:26

  message HasChunksRequest

  chunkstore.proto:69

  • optional RepoId repo_id = 1

  • repeated bytes hashes = 2

  • string repo_token = 3

  • string repo_path = 4

  message HasChunksResponse

  chunkstore.proto:77

  • repeated int32 absent = 1

  • string repo_token = 2

• rpc GetDownloadLocations (GetDownloadLocsRequest, GetDownloadLocsResponse)

  chunkstore.proto:30

  Get the download locations for a list of chunk hashes. Deprecated. Use StreamDownloadLocations.

• rpc StreamDownloadLocations (stream GetDownloadLocsRequest, stream GetDownloadLocsResponse)

  chunkstore.proto:35

  Get the download locations for a list of chunk hashes. Streaming to support large and incrementally available payloads. Results are generated as requests come in.

• rpc GetUploadLocations (GetUploadLocsRequest, GetUploadLocsResponse)

  chunkstore.proto:39

  Get upload locations for a list of table file hashes. NOTE: We upload full table files but download individual chunks.

  message GetUploadLocsRequest

  chunkstore.proto:148

  • optional RepoId repo_id = 1

  • repeated bytes table_file_hashes = 2

  • repeated TableFileDetails table_file_details = 3

  • string repo_token = 4

  • string repo_path = 5

  message GetUploadLocsResponse

  chunkstore.proto:157

  • repeated UploadLoc locs = 1

  • string repo_token = 2

• rpc Rebase (RebaseRequest, RebaseResponse)

  chunkstore.proto:41

  message RebaseRequest

  chunkstore.proto:163

  • optional RepoId repo_id = 1

  • string repo_token = 2

  • string repo_path = 3

  message RebaseResponse

  chunkstore.proto:169

  • string repo_token = 1

• rpc Root (RootRequest, RootResponse)

  chunkstore.proto:43

  message RootRequest

  chunkstore.proto:173

  • optional RepoId repo_id = 1

  • string repo_token = 2

  • string repo_path = 3

  message RootResponse

  chunkstore.proto:180

  • bytes root_hash = 1

  • string repo_token = 2

• rpc Commit (CommitRequest, CommitResponse)

  chunkstore.proto:45

  message CommitRequest

  chunkstore.proto:191

  • optional RepoId repo_id = 1

  • bytes current = 2

  • bytes last = 3

  • repeated ChunkTableInfo chunk_table_info = 4

  • optional ClientRepoFormat client_repo_format = 14

  • string repo_path = 5

  message CommitResponse

  chunkstore.proto:200

  • bool success = 1

• rpc ListTableFiles (ListTableFilesRequest, ListTableFilesResponse)

  chunkstore.proto:47

  message ListTableFilesRequest

  chunkstore.proto:256

  • optional RepoId repo_id = 1

  • bool appendix_only = 2

  • string repo_token = 3

  • string repo_path = 4

  message ListTableFilesResponse

  chunkstore.proto:287

  • bytes root_hash = 1

  • repeated TableFileInfo table_file_info = 2

  • repeated TableFileInfo appendix_table_file_info = 3

  • string repo_token = 4

• rpc RefreshTableFileUrl (RefreshTableFileUrlRequest, RefreshTableFileUrlResponse)

  chunkstore.proto:49

  message RefreshTableFileUrlResponse

  chunkstore.proto:280

  • string url = 1

  • optional google.protobuf.Timestamp refresh_after = 2

  • string repo_token = 3

• rpc AddTableFiles (AddTableFilesRequest, AddTableFilesResponse)

  chunkstore.proto:51

  message AddTableFilesRequest

  chunkstore.proto:301

  • optional RepoId repo_id = 1

  • optional ClientRepoFormat client_repo_format = 2

  • repeated ChunkTableInfo chunk_table_info = 3

  • ManifestAppendixOption appendix_option = 4

    If set, this is a write for the manifest appendix, not just the manifest table file specs. The table files appearing in `chunk_table_info` are added to `specs` and are also set in the manifest appendix. If `appendix_option` is `SET`, then the value of the appendix becomes the full list provided in `chunk_table_info` and any prior specs in the appendix are removed from the manifest specs. If `append_option` is `APPEND`, then the supplied table files are added to the appendix and to specs.

  • string repo_token = 5

  • string repo_path = 6

  message AddTableFilesResponse

  chunkstore.proto:317

  • bool success = 1

  • string repo_token = 2

• rpc WhoAmI (WhoAmIRequest, WhoAmIResponse)

  credentials.proto:22

  message WhoAmIRequest

  credentials.proto:25

  (message has no fields)

  message WhoAmIResponse

  credentials.proto:28

  • string username = 1

    Ex: "bheni"

  • string display_name = 2

    Ex: "Brian Hendriks"

  • string email_address = 3

    Ex: "brian@liquidata.co"

Used in: AddTableFilesRequest, CommitRequest

• bytes hash = 1

• uint32 chunk_count = 2

Used in: AddTableFilesRequest, CommitRequest, GetRepoMetadataRequest

• string nbf_version = 1

• string nbs_version = 2

Used in: GetDownloadLocsResponse

• oneof location

  • HttpGetChunk http_get = 1

  • HttpGetRange http_get_range = 2

• optional google.protobuf.Timestamp refresh_after = 3

• optional RefreshTableFileUrlRequest refresh_request = 4

Used as request type in: ChunkStoreService.GetDownloadLocations, ChunkStoreService.StreamDownloadLocations

• optional RepoId repo_id = 1

• repeated bytes chunk_hashes = 2

• string repo_token = 3

• string repo_path = 4

Used as response type in: ChunkStoreService.GetDownloadLocations, ChunkStoreService.StreamDownloadLocations

• repeated DownloadLoc locs = 1

• string repo_token = 2

Used in: DownloadLoc

• string url = 1

• repeated bytes hashes = 2

Used in: DownloadLoc

• string url = 1

• repeated RangeChunk ranges = 2

Used in: UploadLoc

• string url = 1

Used in: AddTableFilesRequest

• MANIFEST_APPENDIX_OPTION_UNSPECIFIED = 0

• MANIFEST_APPENDIX_OPTION_SET = 1

• MANIFEST_APPENDIX_OPTION_APPEND = 2

A ChunkStore can request a client to implement a specific concurrency control mechanism when updating a branch HEAD. This exists because passive remotes, like DoltHub, typically do not have meaningful workingSets. When a client requests to push a branch HEAD to a DoltHub remote, they have no visibility into the workingSet/ value for the corresponding branch. It has historically been the case that clients ignore it and just push the branch HEAD. On the other hand, when pushing to a running sql-server, not stomping concurrent transaction is important, and the remote endpoint will want the pushing client to ensure that it both checks that the branch's working set is clean and that it updates the branch's working set appropriately if the push is successful. Servers advertise which concurrency control mechanism they want in their GetRepoMetadataResponse.

Used in: GetRepoMetadataResponse

• PUSH_CONCURRENCY_CONTROL_UNSPECIFIED = 0

• PUSH_CONCURRENCY_CONTROL_IGNORE_WORKING_SET = 1

• PUSH_CONCURRENCY_CONTROL_ASSERT_WORKING_SET = 2

Used in: HttpGetRange

• bytes hash = 1

• uint64 offset = 2

• uint32 length = 3

• uint64 dictionary_offset = 4

  Archive zStd dictionary spans are indicated with these fields. On the consumer side, the existence of a non-zero dictionary length indicates that the chunk is compressed using zStd. This implies that the chunkSource is an archive source. If the dictionary length is zero, the chunk is compressed using snappy compression (classic NOMS table file or journal chunk use snappy compression).

• uint32 dictionary_length = 5

Used as request type in: ChunkStoreService.RefreshTableFileUrl

Used as field type in: DownloadLoc, TableFileInfo

• optional RepoId repo_id = 1

• string file_id = 2

• string repo_token = 3

• string repo_path = 4

RepoId is how repositories are represented on dolthub, for example `dolthub/us-housing-prices-v2` has org = dolthub, repo_name = us-housing-prices-v2. All methods of ChunkStoreService targeting a repository take a RepoId. Many will also take a `repo_token`, which can be provided in addition to a RepoId. `repo_token`s can be returned from repository-accessing RPCs on ChunkStoreService, and providing them instead of RepoId for future calls to ChunkStoreService can be more efficient than providing RepoId itself. `repo_token`s are fully opaque to the client.

Used in: AddTableFilesRequest, CommitRequest, GetDownloadLocsRequest, GetRepoMetadataRequest, GetUploadLocsRequest, HasChunksRequest, ListTableFilesRequest, RebaseRequest, RefreshTableFileUrlRequest, RootRequest

• string org = 1

• string repo_name = 2

Used in: GetUploadLocsRequest

• bytes id = 1

• uint64 content_length = 2

• bytes content_hash = 3

• uint64 num_chunks = 4

• string suffix = 5

Used in: ListTableFilesResponse

• string file_id = 1

• uint32 num_chunks = 2

• string url = 3

• optional google.protobuf.Timestamp refresh_after = 4

• optional RefreshTableFileUrlRequest refresh_request = 5

Used in: GetUploadLocsResponse

• bytes table_file_hash = 1

• oneof location

  • HttpPostTableFile http_post = 2