Get desktop application:
View/edit binary Protocol Buffers messages
An operation to update a Counter, either on its own or inside a Map. The `increment` field can be positive or negative. When absent, the meaning is an increment by 1.
Used in: ,
The equivalent of KV's "RpbGetReq", results in a DtFetchResp. The request-time options are limited to ones that are relevant to structured data-types.
The identifier: bucket, key and bucket-type
Request options
Experimental, may change/disappear
Experimental, may change/disappear
For read-only requests or context-free operations, you can set this to false to reduce the size of the response payload.
The response to a "Fetch" request. If the `include_context` option is specified, an opaque "context" value will be returned along with the user-friendly data. When sending an "Update" request, the client should send this context as well, similar to how one would send a vclock for KV updates. The `type` field indicates which value type to expect. When the `value` field is missing from the message, the client should interpret it as a "not found".
Used in:
A "union" type for update operations. The included operation depends on the datatype being updated.
Used in:
Adding values to a hyperloglog (set) is just like adding values to a set.
The equivalent of KV's "RpbPutReq", results in an empty response or "DtUpdateResp" if `return_body` is specified, or the key is assigned by the server. The request-time options are limited to ones that are relevant to structured data-types.
The identifier
missing key results in server-assigned key, like KV
bucket type, not data-type (but the data-type is constrained per bucket-type)
Opaque update-context
The operations
Request options
Experimental, may change/disappear
Experimental, may change/disappear
When return_body is true, should the context be returned too?
The equivalent of KV's "RpbPutResp", contains the assigned key if it was assigned by the server, and the resulting value and context if return_body was set.
The key, if assigned by the server
The opaque update context and value, if return_body was set.
The value of the fetched data type. If present in the response, then empty values (sets, maps) should be treated as such.
Used in:
We return an estimated cardinality of the Hyperloglog set on fetch.
An operation to update a GSet, on its own. GSet members are opaque binary values, you can only add them to a Set.
Used in:
An operation to update a Hyperloglog Set, a top-level DT. You can only add to a HllSet.
Used in:
An entry in a map is a pair of a field-name and value. The type defined in the field determines which value type is expected.
Used in: ,
Field names in maps are composed of a binary identifier and a type. This is so that two clients can create fields with the same name but different types, and they converge independently.
Used in: , ,
The types that can be stored in a map are limited to counters, sets, registers, flags, and maps.
Used in:
An operation to update a Map. All operations apply to individual fields in the Map.
Used in: ,
REMOVE removes a field and value from the Map. UPDATE applies type-specific operations to the values stored in the Map.
An operation to be applied to a value stored in a Map -- the contents of an UPDATE operation. The operation field that is present depends on the type of the field to which it is applied.
Used in:
There is only one operation on a register, which is to set its value, therefore the "operation" is the new value.
Flags only exist inside Maps and can only be enabled or disabled, and there are no arguments to the operations.
Used in:
Authentication request
Preflist item
Used in:
Bucket properties
Used in: , ,
Declared in riak_core_app
Declared in riak_kv_app
Used by riak_kv_multi_backend
Used by riak_search bucket fixup
Search index
KV Datatypes
KV strong consistency
KV fast path
Hyperlolog DT Precision
KV sweeper object expiry time
Used by riak_repl bucket fixup
Used in:
added solely for riak_cs currently for folding over a bucket and returning objects.
Bucket type, if not set we assume the 'default' type
parallel extraction extension
chopped up coverage plan per-req
return for CS bucket fold
A commit hook, which may either be a modfun or a JavaScript named function
Used in:
Content message included in get/put responses Holds the value and associated metadata
Used in: , ,
the media type/format
links to other resources
user metadata stored with the object
user metadata stored with the object
counter value
Counter value response
Counter update request
Counter update response? No message | error response
Segment of a coverage plan
Used in:
Some human readable description of the keyspace covered
Opaque context to pass into 2I query.
Request a segmented coverage plan for the specified bucket
Bucket type, if not set we assume the 'default' type
If undefined, we build a normal coverage plan. If <ring_size, we build a coverage plan with exactly ring_size entries, anything larger will have a power of 2 entries covering keyspaces smaller than a partition
For failure recovery
Other coverage contexts that have failed to assist Riak in deciding what nodes to avoid
Segmented coverage plan response
Delete request
Experimental, may change/disappear
Experimental, may change/disappear
Bucket type, if not set we assume the 'default' type
Error response - may be generated for any Req
Get bucket-key preflist request
Get bucket-key preflist response
Get bucket properties request
Get bucket properties response
Get bucket properties request
Get ClientId Request - no message defined, just send RpbGetClientIdReq message code
Client id in use for this connection
Get Request - retrieve bucket/key
fail if the supplied vclock does not match
return everything but the value
return the tombstone's vclock, if applicable
Experimental, may change/disappear
Experimental, may change/disappear
Bucket type, if not set we assume the 'default' type
Get Response - if the record was not found there will be no content/vclock
Used in:
the opaque vector clock for the object
Get server info request - no message defined, just send RpbGetServerInfoReq message code
Stolen from CS bucket response, to be used when return_body=true
Used in: ,
Secondary Index query request
key here means equals value for index?
Bucket type, if not set we assume the 'default' type
Whether to use pagination sort for non-paginated queries
parallel extraction extension
chopped up coverage plan per-req
Return values with keys, only works with $bucket/$key index queries
Used in:
Secondary Index query response
Link metadata
Used in:
List buckets request
Bucket type, if not set we assume the 'default' type
List buckets response - one or more of these packets will be sent the last one will have done set true (and may not have any buckets in it)
List keys in bucket request
Bucket type, if not set we assume the 'default' type
List keys in bucket response - one or more of these packets will be sent the last one will have done set true (and may not have any keys in it)
Map/Reduce request
Map/Reduce response one or more of these packets will be sent the last one will have done set true (and may not have phase/data in it)
Module-Function pairs for commit hooks and other bucket properties that take functions
Used in: ,
Key/value pair - used for user metadata, indexes, search doc fields
Used in: , , ,
Put request - if options.return_body is set then the updated metadata/data for the key will be returned.
Experimental, may change/disappear
Experimental, may change/disappear
Bucket type, if not set we assume the 'default' type
Put response - same as get response with optional key if one was generated
the opaque vector clock for the object
the key generated, if any
Reset bucket properties request
Used in:
Query string
Index
Limit rows
Starting offset
Sort order
Inline fields filtering query
Default field
Default op
Return fields limit (for ids only, generally)
Presort (key / score)
Result documents
Maximum score
Number of results
Set bucket properties request
Set bucket properties request
Client id to use for this connection
Used in: ,
Index name
Schema name
N value
DELETE request - Remove an index
Index name
GET request - If a name is given, return matching index, else return all
Index name
PUT request - Create a new index
Timeout value
Used in: ,
Index name
Schema data
GET request - Return matching schema by name
Schema name
PUT request - create or potentially update a new schema
An operation to update a Set, either on its own or inside a Map. Set members are opaque binary values, you can only add or remove them from a Set.
Used in: ,
Used in: , ,
Used in: , ,
Used in:
Segment of a TS coverage plan
Used in:
Opaque context to pass into follow-up request
Might be other types of coverage queries/responses
Request a segmented coverage plan for this query
left optional to support parameterized queries in the future
For failure recovery
Other coverage contexts that have failed to assist Riak in deciding what nodes to avoid
Segmented TS coverage plan response
(message has no fields)
0 or 1 rows
Used in: ,
optional: omitting it should use table order
(message has no fields)
Dispatch a query to Riak
left optional to support parameterized queries in the future
chopped up coverage plan per-req
0 to n rows
Each prospective subquery has a range of valid time values
Used in:
Some human readable description of the time range
Used in: , , ,