Documentation ¶
Overview ¶
Example (Pfs) ¶
package main import ( "bytes" "strings" "github.com/pachyderm/pachyderm/src/client" ) func main() { c, err := client.NewFromAddress("0.0.0.0:30650") if err != nil { return // handle error } // Create a repo called "repo" if err := c.CreateRepo("repo"); err != nil { return // handle error } // Start a commit in our new repo on the "master" branch commit1, err := c.StartCommit("repo", "master") if err != nil { return // handle error } // Put a file called "file" in the newly created commit with the content "foo\n". if _, err := c.PutFile("repo", commit1.ID, "file", strings.NewReader("foo\n")); err != nil { return // handle error } // Finish the commit. if err := c.FinishCommit("repo", commit1.ID); err != nil { return //handle error } // Read what we wrote. var buffer bytes.Buffer if err := c.GetFile("repo", commit1.ID, "file", 0, 0, "", false, nil, &buffer); err != nil { return //handle error } // buffer now contains "foo\n" // Start another commit with the previous commit as the parent. commit2, err := c.StartCommit("repo", commit1.ID) if err != nil { return //handle error } // Extend "file" in the newly created commit with the content "bar\n". if _, err := c.PutFile("repo", commit2.ID, "file", strings.NewReader("bar\n")); err != nil { return // handle error } // Finish the commit. if err := c.FinishCommit("repo", commit2.ID); err != nil { return //handle error } // Read what we wrote. buffer.Reset() if err := c.GetFile("repo", commit2.ID, "file", 0, 0, "", false, nil, &buffer); err != nil { return //handle error } // buffer now contains "foo\nbar\n" // We can still read the old version of the file though: buffer.Reset() if err := c.GetFile("repo", commit1.ID, "file", 0, 0, "", false, nil, &buffer); err != nil { return //handle error } // buffer now contains "foo\n" // We can also see the Diff between the most recent commit and the first one: buffer.Reset() if err := c.GetFile("repo", commit2.ID, "file", 0, 0, commit1.ID, false, nil, &buffer); err != nil { return //handle error } }
Output:
Example (Pps) ¶
package main import ( "bytes" "github.com/pachyderm/pachyderm/src/client" "github.com/pachyderm/pachyderm/src/client/pps" ) func main() { c, err := client.NewFromAddress("0.0.0.0:30650") if err != nil { return // handle error } // we assume there's already a repo called "repo" // and that it already has some data in it // take a look at src/client/pfs_test.go for an example of how to get there. // Create a map pipeline if err := c.CreatePipeline( "map", // the name of the pipeline "pachyderm/test_image", // your docker image []string{"map"}, // the command run in your docker image nil, // no stdin nil, // let pachyderm decide the parallelism []*pps.PipelineInput{ // map over "repo" client.NewPipelineInput("repo", client.MapMethod), }, false, // not an update ); err != nil { return // handle error } if err := c.CreatePipeline( "reduce", // the name of the pipeline "pachyderm/test_image", // your docker image []string{"reduce"}, // the command run in your docker image nil, // no stdin nil, // let pachyderm decide the parallelism []*pps.PipelineInput{ // reduce over "map" client.NewPipelineInput("map", client.ReduceMethod), }, false, // not an update ); err != nil { return // handle error } commits, err := c.ListCommitByRepo( // List commits that are... []string{"reduce"}, // from the "reduce" repo (which the "reduce" pipeline outputs) nil, // no provenance client.CommitTypeRead, // are readable client.CommitStatusNormal, // ignore cancelled commits true, // block until commits are available ) if err != nil { return // handle error } for _, commitInfo := range commits { // Read output from the pipeline var buffer bytes.Buffer if err := c.GetFile("reduce", commitInfo.Commit.ID, "file", 0, 0, "", false, nil, &buffer); err != nil { return //handle error } } }
Output:
Index ¶
- Constants
- Variables
- func NewBlock(hash string) *pfs.Block
- func NewCommit(repoName string, commitID string) *pfs.Commit
- func NewFile(repoName string, commitID string, path string) *pfs.File
- func NewJob(jobID string) *pps.Job
- func NewJobInput(repoName string, commitID string, method *pps.Method) *pps.JobInput
- func NewPipeline(pipelineName string) *pps.Pipeline
- func NewPipelineInput(repoName string, method *pps.Method) *pps.PipelineInput
- func NewRepo(repoName string) *pfs.Repo
- type APIClient
- func NewFromAddress(addr string) (*APIClient, error)
- func NewFromAddressWithConcurrency(addr string, maxConcurrentStreams uint) (*APIClient, error)
- func NewInCluster() (*APIClient, error)
- func NewMetricsClientFromAddress(addr string, metrics bool, prefix string) (*APIClient, error)
- func NewMetricsClientFromAddressWithConcurrency(addr string, metrics bool, prefix string, maxConcurrentStreams uint) (*APIClient, error)
- func (c APIClient) ArchiveAll() error
- func (c APIClient) ArchiveCommit(repoName string, commitID string) error
- func (c APIClient) CancelCommit(repoName string, commitID string) error
- func (c *APIClient) Close() error
- func (c APIClient) Compact() error
- func (c APIClient) CreateJob(image string, cmd []string, stdin []string, ...) (*pps.Job, error)
- func (c APIClient) CreatePipeline(name string, image string, cmd []string, stdin []string, ...) error
- func (c APIClient) CreateRepo(repoName string) error
- func (c APIClient) DeleteAll() error
- func (c APIClient) DeleteBlock(block *pfs.Block) error
- func (c APIClient) DeleteCommit(repoName string, commitID string) error
- func (c APIClient) DeleteFile(repoName string, commitID string, path string) error
- func (c APIClient) DeleteJob(jobID string) error
- func (c APIClient) DeletePipeline(name string) error
- func (c APIClient) DeleteRepo(repoName string, force bool) error
- func (c APIClient) FinishCommit(repoName string, commitID string) error
- func (c APIClient) FlushCommit(commits []*pfs.Commit, toRepos []*pfs.Repo) ([]*pfs.CommitInfo, error)
- func (c APIClient) ForkCommit(repoName string, parentCommit string, branch string) (*pfs.Commit, error)
- func (c APIClient) GetBlock(hash string, offset uint64, size uint64) (io.Reader, error)
- func (c APIClient) GetFile(repoName string, commitID string, path string, offset int64, size int64, ...) error
- func (c APIClient) GetLogs(jobID string, writer io.Writer) error
- func (c APIClient) GetObject(hash string, writer io.Writer) error
- func (c APIClient) GetTag(tag string, writer io.Writer) error
- func (c APIClient) InspectBlock(hash string) (*pfs.BlockInfo, error)
- func (c APIClient) InspectCommit(repoName string, commitID string) (*pfs.CommitInfo, error)
- func (c APIClient) InspectFile(repoName string, commitID string, path string, fromCommitID string, ...) (*pfs.FileInfo, error)
- func (c APIClient) InspectJob(jobID string, blockState bool) (*pps.JobInfo, error)
- func (c APIClient) InspectObject(hash string) (*pfs.ObjectInfo, error)
- func (c APIClient) InspectPipeline(pipelineName string) (*pps.PipelineInfo, error)
- func (c APIClient) InspectRepo(repoName string) (*pfs.RepoInfo, error)
- func (c *APIClient) KeepConnected(cancel chan bool)
- func (c APIClient) ListBlock() ([]*pfs.BlockInfo, error)
- func (c APIClient) ListBranch(repoName string, status pfs.CommitStatus) ([]string, error)
- func (c APIClient) ListCommit(exclude []*pfs.Commit, include []*pfs.Commit, provenance []*pfs.Commit, ...) ([]*pfs.CommitInfo, error)
- func (c APIClient) ListCommitByRepo(repoNames []string, provenance []*pfs.Commit, commitType pfs.CommitType, ...) ([]*pfs.CommitInfo, error)
- func (c APIClient) ListFile(repoName string, commitID string, path string, fromCommitID string, ...) ([]*pfs.FileInfo, error)
- func (c APIClient) ListFileFast(repoName string, commitID string, path string, fromCommitID string, ...) ([]*pfs.FileInfo, error)
- func (c APIClient) ListJob(pipelineName string, inputCommit []*pfs.Commit) ([]*pps.JobInfo, error)
- func (c APIClient) ListPipeline() ([]*pps.PipelineInfo, error)
- func (c APIClient) ListRepo(provenance []string) ([]*pfs.RepoInfo, error)
- func (c APIClient) MakeDirectory(repoName string, commitID string, path string) (retErr error)
- func (c APIClient) PutBlock(delimiter pfs.Delimiter, reader io.Reader) (blockRefs *pfs.BlockRefs, retErr error)
- func (c APIClient) PutFile(repoName string, commitID string, path string, reader io.Reader) (_ int, retErr error)
- func (c APIClient) PutFileURL(repoName string, commitID string, path string, url string, recursive bool) (retErr error)
- func (c APIClient) PutFileWithDelimiter(repoName string, commitID string, path string, delimiter pfs.Delimiter, ...) (_ int, retErr error)
- func (c APIClient) PutFileWriter(repoName string, commitID string, path string, delimiter pfs.Delimiter) (io.WriteCloser, error)
- func (c APIClient) PutObject(r io.Reader, tags ...string) (object *pfs.Object, retErr error)
- func (c APIClient) ReadObject(hash string) ([]byte, error)
- func (c APIClient) ReadTag(tag string) ([]byte, error)
- func (c APIClient) ReplayCommit(repo string, fromCommits []string, to string) ([]*pfs.Commit, error)
- func (c APIClient) RerunPipeline(name string, include []*pfs.Commit, exclude []*pfs.Commit) error
- func (c APIClient) SetMaxConcurrentStreams(n int)
- func (c APIClient) SquashCommit(repo string, fromCommits []string, to string) error
- func (c APIClient) StartCommit(repoName string, parentCommit string) (*pfs.Commit, error)
- func (c APIClient) StartPipeline(name string) error
- func (c APIClient) StopPipeline(name string) error
- func (c APIClient) TagObject(hash string, tags ...string) error
- func (c APIClient) Walk(repoName string, commitID string, path string, fromCommitID string, ...) error
- type BlockAPIClient
- type ObjectAPIClient
- type PfsAPIClient
- type PpsAPIClient
- type WalkFn
Examples ¶
Constants ¶
const ( CommitTypeNone = pfs.CommitType_COMMIT_TYPE_NONE CommitTypeRead = pfs.CommitType_COMMIT_TYPE_READ CommitTypeWrite = pfs.CommitType_COMMIT_TYPE_WRITE )
CommitTypes alias pfs.CommitType_*
const ( CommitStatusNormal = pfs.CommitStatus_NORMAL CommitStatusArchived = pfs.CommitStatus_ARCHIVED CommitStatusCancelled = pfs.CommitStatus_CANCELLED CommitStatusAll = pfs.CommitStatus_ALL )
CommitStatus alias pfs.CommitStatus_*
const ( // PPSPodNameEnv is the environment variable that a pod can use to // see its own name. The pod name is made available through the Kubernetes // downward API. PPSPodNameEnv = "PPS_POD_NAME" // PPSLeasePeriodSecsEnv is the amount of time for a lease on a chunk // to expire. // That is, a pod needs to send ContinueJob to PPS at lease once every this // amount of time in order to keep owning a chunk. In reality, pods send // ContinueJob more often than that because they need to account for network // latency. PPSLeasePeriodSecsEnv = "PPS_LEASE_PERIOD_SECS" // PPSHeartbeatSecsEnv controls how many seconds before a pod sends // a heartbeat again. PPSHeartbeatSecsEnv = "PPS_HEARTBEAT_SECS" // PPSMaxHeartbeatRetriesEnv controls how many times a pod can fail // to send heartbeats before it decides to shut itself down. PPSMaxHeartbeatRetriesEnv = "PPS_MAX_HEARTBEAT_RETRIES" )
const DefaultMaxConcurrentStreams uint = 100
DefaultMaxConcurrentStreams defines the max number of Putfiles or Getfiles happening simultaneously
Variables ¶
var ( // MapMethod defines a pps.Method for mapper pipelines. MapMethod = &pps.Method{ Partition: pps.Partition_BLOCK, Incremental: pps.Incremental_DIFF, } // ReduceMethod defines a pps.Method for non-incremental reducer pipelines. ReduceMethod = &pps.Method{ Partition: pps.Partition_FILE, Incremental: pps.Incremental_NONE, } // IncrementalReduceMethod defines a pps.Method for incremental reducer pipelines. IncrementalReduceMethod = &pps.Method{ Partition: pps.Partition_FILE, Incremental: pps.Incremental_DIFF, } // GlobalMethod defines a pps.Method for non-incremental, non-partitioned pipelines. GlobalMethod = &pps.Method{ Partition: pps.Partition_REPO, Incremental: pps.Incremental_NONE, } // DefaultMethod defines the default pps.Method for a pipeline. DefaultMethod = MapMethod // MethodAliasMap maps a string to a pps.Method for JSON decoding. MethodAliasMap = map[string]*pps.Method{ "map": MapMethod, "reduce": ReduceMethod, "incremental_reduce": IncrementalReduceMethod, "global": GlobalMethod, } // ReservedRepoNames defines a set of reserved repo names for internal use. ReservedRepoNames = map[string]bool{ "out": true, "prev": true, } )
var (
// MaxMsgSize is used to define the GRPC frame size
MaxMsgSize = 20 * 1024 * 1024
)
Functions ¶
func NewJobInput ¶
NewJobInput creates a pps.JobInput.
func NewPipeline ¶
NewPipeline creates a pps.Pipeline.
func NewPipelineInput ¶
func NewPipelineInput(repoName string, method *pps.Method) *pps.PipelineInput
NewPipelineInput creates a new pps.PipelineInput
Types ¶
type APIClient ¶
type APIClient struct { PfsAPIClient PpsAPIClient BlockAPIClient ObjectAPIClient // contains filtered or unexported fields }
An APIClient is a wrapper around pfs, pps and block APIClients.
func NewFromAddress ¶
NewFromAddress constructs a new APIClient for the server at addr.
func NewFromAddressWithConcurrency ¶ added in v1.3.6
NewFromAddressWithConcurrency constructs a new APIClient and sets the max concurrency of streaming requests (GetFile / PutFile)
func NewInCluster ¶
NewInCluster constructs a new APIClient using env vars that Kubernetes creates. This should be used to access Pachyderm from within a Kubernetes cluster with Pachyderm running on it.
func NewMetricsClientFromAddress ¶ added in v1.2.4
NewMetricsClientFromAddress Creates a client that will report a user's Metrics
func NewMetricsClientFromAddressWithConcurrency ¶ added in v1.3.6
func NewMetricsClientFromAddressWithConcurrency(addr string, metrics bool, prefix string, maxConcurrentStreams uint) (*APIClient, error)
NewMetricsClientFromAddressWithConcurrency Creates a client that will report a user's Metrics, and sets the max concurrency of streaming requests (GetFile / PutFile)
func (APIClient) ArchiveAll ¶ added in v1.2.0
ArchiveAll archives all commits in all repos.
func (APIClient) ArchiveCommit ¶ added in v1.2.0
ArchiveCommit marks a commit as archived. Archived commits are not listed in ListCommit unless commit status is set to Archived or All. Archived commits are not considered by FlushCommit either.
func (APIClient) CancelCommit ¶
CancelCommit ends the process of committing data to a repo. It differs from FinishCommit in that the Commit will not be used as a source for downstream pipelines. CancelCommit is used primarily by PPS for the output commits of errant jobs.
func (APIClient) CreateJob ¶
func (c APIClient) CreateJob( image string, cmd []string, stdin []string, parallelismSpec *pps.ParallelismSpec, inputs []*pps.JobInput, parentJobID string, internalPort int32, externalPort int32, ) (*pps.Job, error)
CreateJob creates and runs a job in PPS. image is the Docker image to run the job in. cmd is the command passed to the Docker run invocation. NOTE as with Docker cmd is not run inside a shell that means that things like wildcard globbing (*), pipes (|) and file redirects (> and >>) will not work. To get that behavior you should have your command be a shell of your choice and pass a shell script to stdin. stdin is a slice of lines that are sent to your command on stdin. Lines need not end in newline characters. parallelism is how many copies of your container should run in parallel. You may pass 0 for parallelism in which case PPS will set the parallelism based on availabe resources. inputs specifies a set of Commits that will be visible to the job during runtime. parentJobID specifies the a job to use as a parent, it may be left empty in which case there is no parent job. If not left empty your job will use the parent Job's output commit as the parent of its output commit.
func (APIClient) CreatePipeline ¶
func (c APIClient) CreatePipeline( name string, image string, cmd []string, stdin []string, parallelismSpec *pps.ParallelismSpec, inputs []*pps.PipelineInput, update bool, ) error
CreatePipeline creates a new pipeline, pipelines are the main computation object in PPS they create a flow of data from a set of input Repos to an output Repo (which has the same name as the pipeline). Whenever new data is committed to one of the input repos the pipelines will create jobs to bring the output Repo up to data. image is the Docker image to run the jobs in. cmd is the command passed to the Docker run invocation. NOTE as with Docker cmd is not run inside a shell that means that things like wildcard globbing (*), pipes (|) and file redirects (> and >>) will not work. To get that behavior you should have your command be a shell of your choice and pass a shell script to stdin. stdin is a slice of lines that are sent to your command on stdin. Lines need not end in newline characters. parallelism is how many copies of your container should run in parallel. You may pass 0 for parallelism in which case PPS will set the parallelism based on availabe resources. inputs specifies a set of Repos that will be visible to the jobs during runtime. commits to these repos will cause the pipeline to create new jobs to process them. update indicates that you want to update an existing pipeline
func (APIClient) CreateRepo ¶
CreateRepo creates a new Repo object in pfs with the given name. Repos are the top level data object in pfs and should be used to store data of a similar type. For example rather than having a single Repo for an entire project you might have seperate Repos for logs, metrics, database dumps etc.
func (APIClient) DeleteAll ¶
DeleteAll deletes everything in the cluster. Use with caution, there is no undo.
func (APIClient) DeleteBlock ¶
DeleteBlock deletes a block from the block store. NOTE: this is lower level function that's used internally and might not be useful to users.
func (APIClient) DeleteCommit ¶
DeleteCommit deletes a commit. Note it is currently not implemented.
func (APIClient) DeleteFile ¶
DeleteFile deletes a file from a Commit. DeleteFile leaves a tombstone in the Commit, assuming the file isn't written to later attempting to get the file from the finished commit will result in not found error. The file will of course remain intact in the Commit's parent.
func (APIClient) DeletePipeline ¶
DeletePipeline deletes a pipeline along with its output Repo.
func (APIClient) DeleteRepo ¶
DeleteRepo deletes a repo and reclaims the storage space it was using. Note that as of 1.0 we do not reclaim the blocks that the Repo was referencing, this is because they may also be referenced by other Repos and deleting them would make those Repos inaccessible. This will be resolved in later versions. If "force" is set to true, the repo will be removed regardless of errors. This argument should be used with care.
func (APIClient) FinishCommit ¶
FinishCommit ends the process of committing data to a Repo and persists the Commit. Once a Commit is finished the data becomes immutable and future attempts to write to it with PutFile will error.
func (APIClient) FlushCommit ¶
func (c APIClient) FlushCommit(commits []*pfs.Commit, toRepos []*pfs.Repo) ([]*pfs.CommitInfo, error)
FlushCommit blocks until all of the commits which have a set of commits as provenance have finished. For commits to be considered they must have all of the specified commits as provenance. This in effect waits for all of the jobs that are triggered by a set of commits to complete. It returns an error if any of the commits it's waiting on are cancelled due to one of the jobs encountering an error during runtime. If toRepos is not nil then only the commits up to and including those repos will be considered, otherwise all repos are considered. Note that it's never necessary to call FlushCommit to run jobs, they'll run no matter what, FlushCommit just allows you to wait for them to complete and see their output once they do.
func (APIClient) ForkCommit ¶ added in v1.2.0
func (c APIClient) ForkCommit(repoName string, parentCommit string, branch string) (*pfs.Commit, error)
ForkCommit is the same as StartCommit except that the commit is created on a new branch.
func (APIClient) GetBlock ¶
GetBlock returns the content of a block using it's hash. offset specifies a number of bytes that should be skipped in the beginning of the block. size limits the total amount of data returned, note you will get fewer bytes than size if you pass a value larger than the size of the block. If size is set to 0 then all of the data will be returned. NOTE: this is lower level function that's used internally and might not be useful to users.
func (APIClient) GetFile ¶
func (c APIClient) GetFile(repoName string, commitID string, path string, offset int64, size int64, fromCommitID string, fullFile bool, shard *pfs.Shard, writer io.Writer) error
GetFile returns the contents of a file at a specific Commit. offset specifies a number of bytes that should be skipped in the beginning of the file. size limits the total amount of data returned, note you will get fewer bytes than size if you pass a value larger than the size of the file. If size is set to 0 then all of the data will be returned. fromCommitID lets you get only the data which was added after this Commit. shard allows you to downsample the data, returning only a subset of the blocks in the file. shard may be left nil in which case the entire file will be returned
func (APIClient) GetObject ¶ added in v1.3.9
GetObject gets an object out of the object store by hash.
func (APIClient) InspectBlock ¶
InspectBlock returns info about a specific Block.
func (APIClient) InspectCommit ¶
InspectCommit returns info about a specific Commit.
func (APIClient) InspectFile ¶
func (c APIClient) InspectFile(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard) (*pfs.FileInfo, error)
InspectFile returns info about a specific file. fromCommitID lets you get only info which was added after this Commit. shard allows you to downsample the data, returning info about only a subset of the blocks in the file. shard may be left nil in which case info about the entire file will be returned
func (APIClient) InspectJob ¶
InspectJob returns info about a specific job. blockOutput will cause the call to block until the job has been assigned an output commit. blockState will cause the call to block until the job reaches a terminal state (failure or success).
func (APIClient) InspectObject ¶ added in v1.3.9
func (c APIClient) InspectObject(hash string) (*pfs.ObjectInfo, error)
InspectObject returns info about an Object.
func (APIClient) InspectPipeline ¶
func (c APIClient) InspectPipeline(pipelineName string) (*pps.PipelineInfo, error)
InspectPipeline returns info about a specific pipeline.
func (APIClient) InspectRepo ¶
InspectRepo returns info about a specific Repo.
func (*APIClient) KeepConnected ¶ added in v1.2.0
KeepConnected periodically health checks the connection and attempts to reconnect if it becomes unhealthy.
func (APIClient) ListBranch ¶
ListBranch lists the active branches on a Repo.
func (APIClient) ListCommit ¶
func (c APIClient) ListCommit(exclude []*pfs.Commit, include []*pfs.Commit, provenance []*pfs.Commit, commitType pfs.CommitType, status pfs.CommitStatus, block bool) ([]*pfs.CommitInfo, error)
ListCommit lists commits.
exclude and include are filters that either include or exclude the ancestors of the given commits. A commit is considered the ancestor of itself. For instance, ListCommit(include("foo/2")) returns commits foo/0, foo/1, and foo/2, if they exist. In contrast, ListCommit(exclude("foo/2")) returns commits that are *not* foo/0, foo/1, or foo/2.
To get all commits on a given branch, simply include a commit whose ID is the branch name: ListCommit(include("foo"))
To get all commits in a repo, use ListCommitByRepo.
To get all commits, simply don't provide include or exclude.
provenance specifies a set of provenance commits, only commits which have ALL of the specified commits as provenance will be returned unless provenance is nil in which case it is ignored.
commitType specifies the type of commit you want returned, normally CommitTypeRead is the most useful option
status specifies the status of commit you want returned. By default, cancelled or archived commits are not returned.
block, when set to true, will cause ListCommit to block until at least 1 new CommitInfo is available. Using fromCommits and block you can get subscription semantics from ListCommit. commitStatus, controls the statuses of the returned commits. The default value `Normal` will filter out archived and cancelled commits.
func (APIClient) ListCommitByRepo ¶ added in v1.2.4
func (c APIClient) ListCommitByRepo(repoNames []string, provenance []*pfs.Commit, commitType pfs.CommitType, status pfs.CommitStatus, block bool) ([]*pfs.CommitInfo, error)
ListCommitByRepo lists commits in the given repos.
repoNames defines a set of Repos to consider commits from, if repoNames is left nil or empty then the result will be empty.
provenance specifies a set of provenance commits, only commits which have ALL of the specified commits as provenance will be returned unless provenance is nil in which case it is ignored.
commitType specifies the type of commit you want returned, normally CommitTypeRead is the most useful option
status specifies the status of commit you want returned. By default, cancelled or archived commits are not returned.
block, when set to true, will cause ListCommit to block until at least 1 new CommitInfo is available. Using repoNames and block you can get subscription semantics from ListCommit. commitStatus, controls the statuses of the returned commits. The default value `Normal` will filter out archived and cancelled commits.
func (APIClient) ListFile ¶
func (c APIClient) ListFile(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard, recurse bool) ([]*pfs.FileInfo, error)
ListFile returns info about all files in a Commit. fromCommitID lets you get only info which was added after this Commit. shard allows you to downsample the data, returning info about only a subset of the blocks in the files or only a subset of files. shard may be left nil in which case info about all the files and all the blocks in those files will be returned. recurse causes ListFile to accurately report the size of data stored in directories, it makes the call more expensive
func (APIClient) ListFileFast ¶ added in v1.2.3
func (c APIClient) ListFileFast(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard) ([]*pfs.FileInfo, error)
ListFileFast is the same as ListFile except that it doesn't compute the sizes of the files. As a result it's faster than ListFile.
func (APIClient) ListJob ¶
ListJob returns info about all jobs. If pipelineName is non empty then only jobs that were started by the named pipeline will be returned If inputCommit is non-nil then only jobs which took the specific commits as inputs will be returned. The order of the inputCommits doesn't matter.
func (APIClient) ListPipeline ¶
func (c APIClient) ListPipeline() ([]*pps.PipelineInfo, error)
ListPipeline returns info about all pipelines.
func (APIClient) ListRepo ¶
ListRepo returns info about all Repos. provenance specifies a set of provenance repos, only repos which have ALL of the specified repos as provenance will be returned unless provenance is nil in which case it is ignored.
func (APIClient) MakeDirectory ¶
MakeDirectory creates a directory in PFS. Note directories are created implicitly by PutFile, so you technically never need this function unless you want to create an empty directory.
func (APIClient) PutBlock ¶
func (c APIClient) PutBlock(delimiter pfs.Delimiter, reader io.Reader) (blockRefs *pfs.BlockRefs, retErr error)
PutBlock takes a reader and splits the data in it into blocks. Blocks are guaranteed to be new line delimited. Blocks are content addressed and are thus identified by hashes of the content. NOTE: this is lower level function that's used internally and might not be useful to users.
func (APIClient) PutFile ¶
func (c APIClient) PutFile(repoName string, commitID string, path string, reader io.Reader) (_ int, retErr error)
PutFile writes a file to PFS from a reader.
func (APIClient) PutFileURL ¶ added in v1.2.0
func (c APIClient) PutFileURL(repoName string, commitID string, path string, url string, recursive bool) (retErr error)
PutFileURL puts a file using the content found at a URL. The URL is sent to the server which performs the request. recursive allow for recursive scraping of some types URLs for example on s3:// urls.
func (APIClient) PutFileWithDelimiter ¶
func (c APIClient) PutFileWithDelimiter(repoName string, commitID string, path string, delimiter pfs.Delimiter, reader io.Reader) (_ int, retErr error)
PutFileWithDelimiter writes a file to PFS from a reader delimiter is used to tell PFS how to break the input into blocks
func (APIClient) PutFileWriter ¶
func (c APIClient) PutFileWriter(repoName string, commitID string, path string, delimiter pfs.Delimiter) (io.WriteCloser, error)
PutFileWriter writes a file to PFS. NOTE: PutFileWriter returns an io.WriteCloser you must call Close on it when you are done writing.
func (APIClient) PutObject ¶ added in v1.3.9
PutObject puts a value into the object store and tags it with 0 or more tags.
func (APIClient) ReadObject ¶ added in v1.3.9
ReadObject gets an object by hash and returns it directly as []byte.
func (APIClient) ReadTag ¶ added in v1.3.9
ReadTag gets an object by tag and returns it directly as []byte.
func (APIClient) ReplayCommit ¶ added in v1.2.0
func (c APIClient) ReplayCommit(repo string, fromCommits []string, to string) ([]*pfs.Commit, error)
ReplayCommit replays a series of commits on top of commit "to". The replayed commits are the ancestors of the commits in "fromCommits", with no duplicates in case of common ancestors.
func (APIClient) RerunPipeline ¶ added in v1.3.6
RerunPipeline reruns a pipeline over a given set of commits. Exclude and include are filters that either include or exclude the ancestors of the given commits. A commit is considered the ancestor of itself. The behavior is the same as that of ListCommit.
func (APIClient) SetMaxConcurrentStreams ¶ added in v1.3.15
SetMaxConcurrentStreams Sets the maximum number of concurrent streams the client can have. It is not safe to call this operations while operations are outstanding.
func (APIClient) SquashCommit ¶ added in v1.2.0
SquashCommit copies the content of `fromCommits` to `to`, which needs to be an open commit.
func (APIClient) StartCommit ¶
StartCommit begins the process of committing data to a Repo. Once started you can write to the Commit with PutFile and when all the data has been written you must finish the Commit with FinishCommit. NOTE, data is not persisted until FinishCommit is called. parentCommit specifies the parent Commit, upon creation the new Commit will appear identical to the parent Commit, data can safely be added to the new commit without affecting the contents of the parent Commit. You may pass "" as parentCommit in which case the new Commit will have no parent and will initially appear empty. branch is a more convenient way to build linear chains of commits. When a commit is started with a non empty branch the value of branch becomes an alias for the created Commit. This enables a more intuitive access pattern. When the commit is started on a branch the previous head of the branch is used as the parent of the commit.
func (APIClient) StartPipeline ¶ added in v1.2.0
StartPipeline restarts a stopped pipeline.
func (APIClient) StopPipeline ¶ added in v1.2.0
StopPipeline prevents a pipeline from processing things, it can be restarted with StartPipeline.
func (APIClient) Walk ¶ added in v1.3.2
func (c APIClient) Walk(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard, walkFn WalkFn) error
Walk walks the pfs filesystem rooted at path. walkFn will be called for each file found under path, this includes both regular files and directories.
type BlockAPIClient ¶
type BlockAPIClient pfs.BlockAPIClient
BlockAPIClient is an alias for pfs.BlockAPIClient.
type ObjectAPIClient ¶ added in v1.3.9
type ObjectAPIClient pfs.ObjectAPIClient
ObjectAPIClient is an alias for pfs.ObjectAPIClient
Directories ¶
Path | Synopsis |
---|---|
Package health is a generated protocol buffer package.
|
Package health is a generated protocol buffer package. |
Package pfs is a generated protocol buffer package.
|
Package pfs is a generated protocol buffer package. |
pkg
|
|
config
Package config is a generated protocol buffer package.
|
Package config is a generated protocol buffer package. |
shard
Package shard is a generated protocol buffer package.
|
Package shard is a generated protocol buffer package. |
Package pps is a generated protocol buffer package.
|
Package pps is a generated protocol buffer package. |
versionpb
Package versionpb is a generated protocol buffer package.
|
Package versionpb is a generated protocol buffer package. |