User Command Line Interface

Slack Docker Pulls

Alluxio’s command line interface provides users with basic file system operations. You can invoke the following command line utility to get all the subcommands:

$ ./bin/alluxio
Usage: alluxio [COMMAND]
       [format [-s]]
       [getConf [key]]
       [logLevel]
       [runTests]
       ...

General operations

This section lists usages and examples of general Alluxio operations with the exception of file system commands which are covered in the Admin CLI doc.

format

The format command formats the Alluxio master and all its workers.

If -s specified, only format if under storage is local and does not already exist

Running this command on an existing Alluxio cluster deletes everything persisted in Alluxio, including cached data and any metadata information. Data in under storage will not be changed.

Warning: format is required when you run Alluxio for the first time. format should only be called while the cluster is not running.

$ ./bin/alluxio format
$ ./bin/alluxio format -s

formatJournal

The formatJournal command formats the Alluxio master journal on this host.

The Alluxio master stores various forms of metadata, including:

  • file system operations
  • where files are located on workers
  • journal transactions
  • under storage file metadata

All this information is deleted if formatJournal is run.,

Warning: formatJournal should only be called while the cluster is not running.

$ ./bin/alluxio formatJournal

formatMasters

The formatMasters command formats the Alluxio masters.

This command defers to formatJournal, but if the UFS is an embedded journal it will format all master nodes listed in the ‘conf/masters’ file instead of just this host.

The Alluxio master stores various forms of metadata, including:

  • file system operations
  • where files are located on workers
  • journal transactions
  • under storage file metadata

All this information is deleted if formatMasters is run.,

Warning: formatMasters should only be called while the cluster is not running.

$ ./bin/alluxio formatMasters

formatWorker

The formatWorker command formats the Alluxio worker on this host.

An Alluxio worker caches files and objects.

formatWorker deletes all the cached data stored in this worker node. Data in under storage will not be changed.

Warning: formatWorker should only be called while the cluster is not running.

$ ./bin/alluxio formatWorker

bootstrapConf

The bootstrapConf command generates the bootstrap configuration file ${ALLUXIO_HOME}/conf/alluxio-site.properties with alluxio.master.hostname set to the passed in value if the configuration file does not exist.

$ ./bin/alluxio bootstrapConf <ALLUXIO_MASTER_HOSTNAME>

Note: This command does not require the Alluxio cluster to be running.

fs

See File System Operations.

fsadmin

The fsadmin command is meant for administrators of the Alluxio cluster. It provides added tools for diagnostics and troubleshooting. For more information see the Admin CLI main page.

Note: This command requires the Alluxio cluster to be running.

getConf

The getConf command prints the configured value for the given key. If the key is invalid, it returns a nonzero exit code. If the key is valid but isn’t set, an empty string is printed. If no key is specified, the full configuration is printed.

Options:

  • --master option prints any configuration properties used by the master.
  • --source option prints the source of the configuration properties.
  • --unit <arg> option displays the configuration value in the given unit. For example, with --unit KB, a configuration value of 4096B returns as 4, and with --unit S, a configuration value of 5000ms returns as 5. Possible unit options include B, KB, MB, GB, TP, PB as units of byte size and MS, S, M, H, D as units of time.
# Displays all the current node configuration
$ ./bin/alluxio getConf

# Displays the value of a property key
$ ./bin/alluxio getConf alluxio.master.hostname

# Displays the configuration of the current running Alluxio leading master
$ ./bin/alluxio getConf --master

# Also display the source of the configuration
$ ./bin/alluxio getConf --source

# Displays the values in a given unit
$ ./bin/alluxio getConf --unit KB alluxio.user.block.size.bytes.default
$ ./bin/alluxio getConf --unit S alluxio.master.journal.flush.timeout

Note: This command does not require the Alluxio cluster to be running.

job

The job command is a tool for interacting with the job service.

The usage is job [generic options] where [generic options] can be one of the following values:

  • leader: Prints the hostname of the job master service leader.
  • ls: Prints the IDs of the most recent jobs, running and finished, in the history up to the capacity set in alluxio.job.master.job.capacity.
  • stat [-v] <id>:Displays the status info for the specific job. Use -v flag to display the status of every task.
  • cancel <id>: Cancels the job with the corresponding id asynchronously.
# Prints the hostname of the job master service leader.
$ ./bin/alluxio job leader

# Prints the IDs, job names, and completion status of the most recently created jobs.
$ ./bin/alluxio job ls
1576539334518 Load COMPLETED
1576539334519 Load CREATED
1576539334520 Load CREATED
1576539334521 Load CREATED
1576539334522 Load CREATED
1576539334523 Load CREATED
1576539334524 Load CREATED
1576539334525 Load CREATED
1576539334526 Load CREATED

# Displays the status info for the specific job.
$ bin/alluxio job stat -v 1579102592778
ID: 1579102592778
Name: Migrate
Description: MigrateConfig{source=/test, destination=/test2, writeType=ASYNC_THROUGH, overwrite=true, delet...
Status: CANCELED
Task 0
Worker: localhost
Status: CANCELED
Task 1
Worker: localhost
Status: CANCELED
Task 2
Worker: localhost
Status: CANCELED
...
...

# Cancels the job asynchronously based on a specific job.
$ bin/alluxio job cancel 1579102592778

$ bin/alluxio job stat 1579102592778 | grep "Status"
Status: CANCELED

Note: This command requires the Alluxio cluster to be running.

logLevel

The logLevel command returns the current value of or updates the log level of a particular class on specific instances. Users are able to change Alluxio server-side log levels at runtime.

The command follows the format alluxio logLevel --logName=NAME [--target=<master|workers|job_master|job_workers|host:port>] [--level=LEVEL], where:

  • --logName <arg> indicates the logger’s class (e.g. alluxio.master.file.DefaultFileSystemMaster)
  • --target <arg> lists the Alluxio master or workers to set. The target could be of the form <master|workers|job_master|job_workers|host:webPort> and multiple targets can be listed as comma-separated entries. The host:webPort format can only be used when referencing a worker. The default target value is the primary master, primary job master, all workers and job workers.
  • --level <arg> If provided, the command changes to the given logger level, otherwise it returns the current logger level.

See here for more examples.

Note: This command requires the Alluxio cluster to be running. You are not able to set the logger level on the standby masters. The standby masters/job masters do not have a running web server. So they are not accepting the requests from this command. If you want to modify the logger level for standby masters, update the log4j.properties and restart the process.

privileges

The privileges command allows an Alluxio cluster administrator to grant, revoke, or deny privileges to a group of users within an Alluxio cluster. See the privileges model documentation for more information.

Note that to enable privileges you’ll need to set alluxio.security.privileges.enabled=true in the Alluxio master’s alluxio-site.properties.

The basic set of privileges is

  • FREE
  • PIN
  • REPLICATION
  • TTL

Examples:

Show privileges for all groups

./bin/alluxio privileges --help # Shows the help text
./bin/alluxio privileges list # lists privileges for all groups

Revoke all privileges

./bin/alluxio privileges revoke -group <user_group> -privileges all # Afterwards, users in this group have no privileges
./bin/alluxio privileges list -group <user_group> # lists all privileges for <user_group>

Grant a single privilege

./bin/alluxio privileges grant -group <user_group> -privileges replication # Afterwards, users in this group can set replication
./bin/alluxio privileges list -group <user_group> # lists all privileges for <user_group>

See privileges model documentation for a more comprehensive set of usages of Alluxio privileges.

runClass

The runClass command runs the main method of an Alluxio class.

For example, to run the multi-mount demo:

$ ./bin/alluxio runClass alluxio.examples.MultiMount <HDFS_URL>

runTest

The runTest command runs end-to-end tests on an Alluxio cluster.

The usage is runTest [--directory <path>] [--operation <operation type>] [--readType <read type>] [--writeType <write type>].

  • --directory Alluxio path for the tests working directory. Default: ${ALLUXIO_HOME}
  • --operation The operation to test, one of BASIC or BASIC_NON_BYTE_BUFFER. By default both operations are tested.
  • --readType The read type to use, one of NO_CACHE, CACHE, CACHE_PROMOTE. By default all readTypes are tested.
  • --writeType The write type to use, one of MUST_CACHE, CACHE_THROUGH, THROUGH, ASYNC_THROUGH. By default all writeTypes are tested.
$ ./bin/alluxio runTest

$ ./bin/alluxio runTest --operation BASIC --readType CACHE --writeType MUST_CACHE

Note: This command requires the Alluxio cluster to be running.

runTests

The runTests command runs all the end-to-end tests on an Alluxio cluster to provide a comprehensive sanity check.

This command is equivalent to running runTest with all the default flag values.

$ ./bin/alluxio runTests

Note: This command requires the Alluxio cluster to be running.

runJournalCrashTest

The runJournalCrashTest simulates a failover to test recovery from the journal.

Note: This command will stop any Alluxio services running on the machine.

runHmsTests

The runHmsTests aims to validate the configuration, connectivity, and permissions of an existing hive metastore which is an important component in compute workflows with Alluxio.

  • -h provides detailed guidance.
  • -m <hive_metastore_uris> (required) the full hive metastore uris to connect to an existing hive metastore.
  • -d <database_name> the database to run tests against. Use default database if not provided.
  • -t [table_name_1,table_name_2,...] tables to run tests against. Run tests against five out of all tables in the given database if not provided.
  • -st <timeout> socket timeout of hive metastore client in minutes.
$ ./bin/alluxio runHmsTests -m thrift://<hms_host>:<hms_port> -d tpcds -t store_sales,web_sales

This tool is suggested to run from compute application environments and checks

  • if the given hive metastore uris are valid
  • if the hive metastore client connection can be established with the target server
  • if hive metastore client operations can be run against the given database and tables

Note: This command does not require the Alluxio cluster to be running.

runHdfsMountTests

The runHdfsMountTests command aims to validate the configuration, connectivity and permissions of an HDFS path. It validates various aspects for connecting to HDFS with the given Alluxio configurations and identifies issues before the path is mounted to Alluxio. This tool will validate a few criteria and return the feedback. If a test failed, advice will be given correspondingly on how the user can rectify the setup.

Usage: runHdfsMountTests [--readonly] [--shared] [--option <key=val>] <hdfsURI>

  • --help provides detailed guidance.
  • --readonly specifies the mount point should be readonly in Alluxio.
  • --shared specifies the mount point should be accessible for all Alluxio users.
  • --option <key>=<val> passes an property to this mount point.
  • <hdfs-path> (required) specifies the HDFS path you want to validate (then mount to Alluxio)

The arguments to this command should be consistent to what you give to the Mount command, in order to validate the setup for the mount.

# If this is your mount command
$ bin/alluxio fs mount --readonly --option alluxio.underfs.version=2.7 \
  --option alluxio.underfs.hdfs.configuration=/etc/hadoop/core-site.xml:/etc/hadoop/hdfs-site.xml \
  <alluxio-path> hdfs://<hdfs-path>

# Pass the same options to runHdfsMountTests
$ bin/alluxio runHdfsMountTests --readonly --option alluxio.underfs.version=2.7 \
  --option alluxio.underfs.hdfs.configuration=/etc/hadoop/core-site.xml:/etc/hadoop/hdfs-site.xml \
  hdfs://<hdfs-path>

Note: This command DOES NOT mount the HDFS path to Alluxio. This command does not require the Alluxio cluster to be running.

runUfsIOTest

The runUfsIOTest command measures the read/write IO throughput from Alluxio cluster to the target HDFS.

Usage: runUfsIOTest --path <hdfs-path> [--io-size <io-size>] [--threads <thread-num>] [--cluster] [--cluster-limit <worker-num>] --java-opt <java-opt>

  • -h, --help provides detailed guidance.
  • --path <hdfs-path> (required) specifies the path to write/read temporary data in.
  • --io-size <io-size> specifies the amount of data each thread writes/reads. It defaults to “4G”.
  • --threads <thread-num> specifies the number of threads to concurrently use on each worker. It defaults to 4.
  • --cluster specifies the benchmark is run in the Alluxio cluster. If not specified, this benchmark will run locally.
  • --cluster-limit <worker-num> specifies how many Alluxio workers to run the benchmark concurrently. If >0, it will only run on that number of workers. If 0, it will run on all available cluster workers. If <0, will run on the workers from the end of the worker list. This flag is only used if --cluster is enabled. This default to 0.
  • --java-opt <java-opt> The java options to add to the command line to for the task. This can be repeated. The options must be quoted and prefixed with a space. For example: --java-opt " -Xmx4g" --java-opt " -Xms2g".

Examples:

# This runs the I/O benchmark to HDFS in your process locally
$ bin/alluxio runUfsIOTest --path hdfs://<hdfs-address>

# This invokes the I/O benchmark to HDFS in the Alluxio cluster
# 1 worker will be used. 4 threads will be created, each writing then reading 4G of data
$ bin/alluxio runUfsIOTest --path hdfs://<hdfs-address> --cluster --cluster-limit 1

# This invokes the I/O benchmark to HDFS in the Alluxio cluster
# 2 workers will be used
# 2 threads will be created on each worker
# Each thread is writing then reading 512m of data
$ bin/alluxio runUfsIOTest --path hdfs://<hdfs-address> --cluster --cluster-limit 2 \
  --io-size 512m --threads 2

Note: This command requires the Alluxio cluster to be running.

runUfsTests

The runUfsTests aims to test the integration between Alluxio and the given UFS. UFS tests validate the semantics Alluxio expects of the UFS.

Usage: runUfsTests --path <ufs_path>

  • --help provides detailed guidance.
  • --path <ufs_path> (required) the full UFS path to run tests against.

The usage of this command includes:

  • Test if the given UFS credentials are valid before mounting the UFS to an Alluxio cluster.
  • If the given UFS is S3, this test can also be used as a S3 compatibility test to test if the target under filesystem can fulfill the minimum S3 compatibility requirements in order to work well with Alluxio.
  • Validate the contract between Alluxio and the given UFS. This is primarily intended for Alluxio developers. Developers are required to add test coverage for changes to an Alluxio UFS module and run those tests to validate.
# Run tests against local UFS
$ ./bin/alluxio runUfsTests --path /local/underfs/path

# Run tests against S3
$ ./bin/alluxio runUfsTests --path s3://<s3_bucket_name> \
  -Ds3a.accessKeyId=<access_key> -Ds3a.secretKey=<secret_key> \
  -Dalluxio.underfs.s3.endpoint=<endpoint_url> -Dalluxio.underfs.s3.disable.dns.buckets=true

Note: This command does not require the Alluxio cluster to be running.

readJournal

The readJournal command parses the current journal and outputs a human readable version to the local folder. Note this command may take a while depending on the size of the journal. Note that Alluxio master is required to stop before reading the local embedded journal.

  • -help provides detailed guidance.
  • -start <arg> the start log sequence number (exclusive). (Default: 0)
  • -end <arg> the end log sequence number (exclusive). (Default: +inf)
  • -inputDir <arg> the input directory on-disk to read journal content from. (Default: Read from system configuration)
  • -outputDir <arg> the output directory to write journal content to. (Default: journal_dump-${timestamp})
  • -master <arg> (advanced) the name of the master (e.g. FileSystemMaster, BlockMaster). (Default: “FileSystemMaster”)
$ ./bin/alluxio readJournal
Dumping journal of type EMBEDDED to /Users/alluxio/journal_dump-1602698211916
2020-10-14 10:56:51,960 INFO  RaftStorageDirectory - Lock on /Users/alluxio/alluxio/journal/raft/02511d47-d67c-49a3-9011-abb3109a44c1/in_use.lock acquired by nodename 78602@alluxio-user
2020-10-14 10:56:52,254 INFO  RaftJournalDumper - Read 223 entries from log /Users/alluxio/alluxio/journal/raft/02511d47-d67c-49a3-9011-abb3109a44c1/current/log_0-222.

Note: This command requires that the Alluxio cluster is NOT running.

upgradeJournal

The upgradeJournal command upgrades an Alluxio journal version 0 (Alluxio version < 1.5.0) to an Alluxio journal version 1 (Alluxio version >= 1.5.0).

-journalDirectoryV0 <arg> will provide the v0 journal persisted location.
It is assumed to be the same as the v1 journal directory if not set.

$ ./bin/alluxio upgradeJournal

Note: This command does not require the Alluxio cluster to be running.

killAll

The killAll command kills all processes containing the specified word.

Note: This kills non-Alluxio processes as well.

copyDir

The copyDir command copies the directory at PATH to all master nodes listed in conf/masters and all worker nodes listed in conf/workers.

$ ./bin/alluxio copyDir conf/alluxio-site.properties

Note: This command does not require the Alluxio cluster to be running.

clearCache

The clearCache command drops the OS buffer cache.

Note: This command does not require the Alluxio cluster to be running.

docGen

The docGen command autogenerates documentation based on the current source code.

Usage: docGen [--metric] [--conf]

  • --metric flag indicates to generate Metric docs
  • --conf flag indicates to generate Configuration docs

Supplying neither flag will default to generating both docs.

Note: This command does not require the Alluxio cluster to be running.

table

See Table Operations.

version

The version command prints Alluxio version.

Usage: version --revision [revision_length]

  • -r,--revision [revision_length] Prints the git revision along with the Alluxio version. Optionally specify the revision length.
$ ./bin/alluxio version

Note: This command does not require the Alluxio cluster to be running.

validateConf

The validateConf command validates the local Alluxio configuration files, checking for common misconfigurations.

$ ./bin/alluxio validateConf

Note: This command does not require the Alluxio cluster to be running.

validateEnv

Before starting Alluxio, it is recommended to ensure that the system environment is compatible with running Alluxio services. The validateEnv command runs checks against the system and reports any potential problems that may prevent Alluxio from starting properly.

The usage is validateEnv COMMAND [NAME] [OPTIONS] where COMMAND can be one of the following values:

  • local: run all validation tasks on the local machine
  • master: run master validation tasks on the local machine
  • worker: run worker validation tasks on the local machine
  • all: run corresponding validation tasks on all master and worker nodes
  • masters: run master validation tasks on all master nodes
  • workers: run worker validation tasks on all worker nodes
  • list: list all validation tasks
# Runs all validation tasks on the local machine
$ ./bin/alluxio validateEnv local

# Runs corresponding validation tasks on all master and worker nodes
$ ./bin/alluxio validateEnv all

# Lists all validation tasks
$ ./bin/alluxio validateEnv list

For all commands except list, NAME specifies the leading prefix of any number of tasks. If NAME is not given, all tasks for the given COMMAND will run.

# Only run validation tasks that check your local system resource limits
$ ./bin/alluxio validateEnv ulimit
# Only run the tasks start with "ma", like "master.rpc.port.available" and "master.web.port.available"
$ ./bin/alluxio validateEnv local ma

OPTIONS can be a list of command line options. Each option has the format -<optionName> [optionValue] For example, [-hadoopConfDir <arg>] could set the path to server-side hadoop configuration directory when running validating tasks.

Note: This command does not require the Alluxio cluster to be running.

collectInfo

The collectInfo command collects information to troubleshoot an Alluxio cluster. For more information see the collectInfo command page.

Note: This command does not require the Alluxio cluster to be running. But if the cluster is not running, this command will fail to gather some information from it.

File System Operations

./bin/alluxio fs
Usage: alluxio fs [generic options]
       [cat <path>]
       [checkConsistency [-r] <Alluxio path>]
       ...

For fs subcommands that take Alluxio URIs as argument (e.g. ls, mkdir), the argument should be either a complete Alluxio URI, such as alluxio://<master-hostname>:<master-port>/<path>, or a path without its header, such as /<path>, to use the default hostname and port set in the conf/allluxio-site.properties.

Note: This command requires the Alluxio cluster to be running.

Wildcard input

Most of the commands which require path components allow wildcard arguments for ease of use. For example:

$ ./bin/alluxio fs rm '/data/2014*'

The example command deletes anything in the data directory with a prefix of 2014.

Note that some shells will attempt to glob the input paths, causing strange errors (Note: the number 21 could be different and comes from the number of matching files in your local filesystem):

rm takes 1 arguments,  not 21

As a workaround, you can disable globbing (depending on the shell type; for example, set -f) or by escaping wildcards, for example:

$ ./bin/alluxio fs cat /\\*

Note the double escape; this is because the shell script will eventually call a java program which should have the final escaped parameters (cat /\\*).

cat

The cat command prints the contents of a file in Alluxio to the console. If you wish to copy the file to your local file system, copyToLocal should be used.

For example, when testing a new computation job, cat can be used as a quick way to check the output:

$ ./bin/alluxio fs cat /output/part-00000

checkConsistency

The checkConsistency command compares Alluxio and under storage metadata for a given path. If the path is a directory, the entire subtree will be compared. The command returns a message listing each inconsistent file or directory. The system administrator should reconcile the differences of these files at their discretion. To avoid metadata inconsistencies between Alluxio and under storages, design your systems to modify files and directories through Alluxio and avoid directly modifying the under storage.

If the -r option is used, the checkConsistency command will repair all inconsistent files and directories under the given path. If an inconsistent file or directory exists only in under storage, its metadata will be added to Alluxio. If an inconsistent file exists in Alluxio and its data is fully present in Alluxio, its metadata will be loaded to Alluxio again.

If the -t <thread count> option is specified, the provided number of threads will be used when repairing consistency. Defaults to the number of CPU cores available,

  • This option has no effect if -r is not specified

NOTE: This command requires a read lock on the subtree being checked, meaning writes and updates to files or directories in the subtree cannot be completed until this command completes.

For example, checkConsistency can be used to periodically validate the integrity of the namespace.

# List each inconsistent file or directory
$ ./bin/alluxio fs checkConsistency /

# Repair the inconsistent files or directories
$ ./bin/alluxio fs checkConsistency -r /

checksum

The checksum command outputs the md5 value of a file in Alluxio.

For example, checksum can be used to verify the contents of a file stored in Alluxio.

$ ./bin/alluxio fs checksum /LICENSE
md5sum: bf0513403ff54711966f39b058e059a3
md5 LICENSE
MD5 (LICENSE) = bf0513403ff54711966f39b058e059a3

chgrp

The chgrp command changes the group of the file or directory in Alluxio. Alluxio supports file authorization with Posix file permission. Group is an authorizable entity in Posix file permissions model. The file owner or super user can execute this command to change the group of the file or directory.

Adding -R option also changes the group of child file and child directory recursively.

For example, chgrp can be used as a quick way to change the group of file:

$ ./bin/alluxio fs chgrp alluxio-group-new /input/file1

chmod

The chmod command changes the permission of file or directory in Alluxio. Currently, octal mode is supported: the numerical format accepts three octal digits which refer to permissions for the file owner, the group and other users. Here is number-permission mapping table:

NumberPermissionrwx
7 read, write and execute rwx
6 read and write rw-
5 read and execute r-x
4 read only r--
3 write and execute -wx
2 write only -w-
1 execute only --x
0 none ---

Adding -R option also changes the permission of child file and child directory recursively.

For example, chmod can be used as a quick way to change the permission of file:

$ ./bin/alluxio fs chmod 755 /input/file1

chown

The chown command changes the owner of the file or directory in Alluxio. For security reasons, the ownership of a file can only be altered by a super user.

For example, chown can be used as a quick way to change the owner of file:

$ ./bin/alluxio fs chown alluxio-user /input/file1
$ ./bin/alluxio fs chown alluxio-user:alluxio-group /input/file2

Adding -R option also changes the owner of child file and child directory recursively.

copyFromLocal

The copyFromLocal command copies the contents of a file in the local file system into Alluxio. If the node you run the command from has an Alluxio worker, the data will be available on that worker. Otherwise, the data will be copied to a random remote node running an Alluxio worker. If a directory is specified, the directory and all its contents will be copied recursively (parallel at file level up to the number of available threads).

Usage: copyFromLocal [--thread <num>] [--buffersize <bytes>] <src> <remoteDst>

  • --thread <num> (optional) Number of threads used to copy files in parallel, default value is CPU cores * 2
  • --buffersize <bytes> (optional) Read buffer size in bytes, default is 8MB when copying from local and 64MB when copying to local
  • <src> file or directory path on the local filesystem
  • <remoteDst> file or directory path on the Alluxio filesystem

For example, copyFromLocal can be used as a quick way to inject data into the system for processing:

$ ./bin/alluxio fs copyFromLocal /local/data /input

copyToLocal

The copyToLocal command copies a file in Alluxio to the local file system. If a directory is specified, the directory and all its contents will be copied recursively.

Usage: copyToLocal [--buffersize <bytes>] <src> <localDst>

  • --buffersize <bytes> (optional) file transfer buffer size in bytes
  • <src> file or directory path on the Alluxio filesystem
  • <localDst> file or directory path on the local filesystem

For example, copyToLocal can be used as a quick way to download output data for additional investigation or debugging.

$ ./bin/alluxio fs copyToLocal /output/part-00000 part-00000
$ wc -l part-00000

count

The count command outputs the number of files and folders matching a prefix as well as the total size of the files. count works recursively and accounts for any nested directories and files.

Usage: count [-h] <dir>

  • -h (optional) print sizes in human readable format (e.g. 1KB 234MB 2GB)
  • <dir> file or directory path in the Alluxio filesystem
$ ./bin/alluxio fs count -h /LICENSE
File Count               Folder Count             Folder Size
1                        0                        26.41KB

count is best utilized when the user has some predefined naming conventions for their files. For example, if data files are stored by their date, count can be used to determine the number of data files and their total size for any date, month, or year.

cp

The cp command copies a file or directory in the Alluxio file system or between the local file system and Alluxio file system.

Scheme file:// indicates the local file system whereas scheme alluxio:// or no scheme indicates the Alluxio file system.

If the -R option is used and the source designates a directory, cp copies the entire subtree at source to the destination.

Usage: cp [--thread <num>] [--buffersize <bytes>] [--preserve] <src> <dst>

  • --thread <num> (optional) Number of threads used to copy files in parallel, default value is CPU cores * 2
  • --buffersize <bytes> (optional) Read buffer size in bytes, default is 8MB when copying from local and 64MB when copying to local
  • --preserve (optional) Preserve file permission attributes when copying files. All ownership, permissions and ACLs will be preserved.
  • <src> source file or directory path
  • <dst> destination file or directory path

For example, cp can be used to copy files between under storage systems.

$ ./bin/alluxio fs cp /hdfs/file1 /s3/

distributedCp

The distributedCp command copies a file or directory in the Alluxio file system distributed across workers using the job service. By default, the command runs synchronously and the user will get a JOB_CONTROL_ID after the command successfully submits the job to be executed. The command will wait until the job is complete, at which point the user will see the list of files copied and statistics on which files completed or failed. The command can also run in async mode with the --async flag. Similar to before, the user will get a JOB_CONTROL_ID after the command successfully submits the job. The difference is that the command will not wait for the job to finish. Users can use the getCmdStatus command with the JOB_CONTROL_ID as an argument to check detailed status information about the job.

If the source designates a directory, distributedCp copies the entire subtree at source to the destination.

Options:

  • --active-jobs: Limits how many jobs can be submitted to the Alluxio job service at the same time. Later jobs must wait until some earlier jobs to finish. The default value is 3000. A lower value means slower execution but also being nicer to the other users of the job service.
  • --overwrite: Whether to overwrite the destination. Default is true.
  • --batch-size: Specifies how many files to be batched into one request. The default value is 20. Notice that if some task failed in the batched job, the whole batched job would fail with some completed tasks and some failed tasks.
  • --async: Specifies whether to wait for command execution to finish. If not explicitly shown then default to run synchronously. 
    $ ./bin/alluxio fs distributedCp --active-jobs 2000 /data/1023 /data/1024
Sample Output:
Please wait for command submission to finish..
Submitted successfully, jobControlId = JOB_CONTROL_ID_1
Waiting for the command to finish ...
Get command status information below:
Successfully copied path /data/1023/$FILE_PATH_1
Successfully copied path /data/1023/$FILE_PATH_2
Successfully copied path /data/1023/$FILE_PATH_3
Total completed file count is 3, failed file count is 0
Finished running the command, jobControlId = JOB_CONTROL_ID_1

# Turn on async submission mode. Run this command to get JOB_CONTROL_ID, then use getCmdStatus to check command detailed status.
$ ./bin/alluxio fs distributedCp /data/1023 /data/1025 --async
Sample Output:
Entering async submission mode.
Please wait for command submission to finish..
Submitted migrate job successfully, jobControlId = JOB_CONTROL_ID_2

Please note below are known limitations for the distributed copy command.

  • Limited Scalability: No more than 1 million total number of files should be moved concurrently. Note that a copy job may stay active for a short period after the last file is copied.
  • Manual Integrity Validation: Verification between source and destination files relies on the response code from the underlying data lake storage. In case the response code is unreliable, we recommend manual verification of source and destination checksums.
  • Manual Cleanup: In certain failure scenarios, a user may need to manually remove partially written contents in destination directories and restart the failed jobs.
  • Limited Observability: Status checks are limited to using the command line for each job individually.

du

The du command outputs the total size and amount stored in Alluxio of files and folders. If a directory is specified, it will display the sizes of all files in this directory.

Usage: du [-s] [-h] [--memory] [-g] <dir>

  • -s (optional) display the aggregate summary of file lengths being displayed
  • -h (optional) print sizes in human readable format (e.g. 1KB 234MB 2GB)
  • -m,--memory (optional) display the in memory size and in memory percentage
  • -g (optional) display information for In-Alluxio data size under the path, grouped by worker
  • <dir> file or directory path in the Alluxio filesystem
# Shows the size information of all the files in root directory
$ ./bin/alluxio fs du /
File Size     In Alluxio       Path
1337          0 (0%)           /alluxio-site.properties
4352          4352 (100%)      /testFolder/NOTICE
26847         0 (0%)           /testDir/LICENSE
2970          2970 (100%)      /testDir/README.md

# Shows the in memory size information
$ ./bin/alluxio fs du --memory /
File Size     In Alluxio       In Memory        Path
1337          0 (0%)           0 (0%)           /alluxio-site.properties
4352          4352 (100%)      4352 (100%)      /testFolder/NOTICE
26847         0 (0%)           0 (0%)           /testDir/LICENSE
2970          2970 (100%)      2970 (100%)      /testDir/README.md

# Shows the aggregate size information in human-readable format
./bin/alluxio fs du -h -s /
File Size     In Alluxio       In Memory        Path
34.67KB       7.15KB (20%)     7.15KB (20%)     /

# Can be used to detect which folders are taking up the most space
./bin/alluxio fs du -h -s /\\*
File Size     In Alluxio       Path
1337B         0B (0%)          /alluxio-site.properties
29.12KB       2970B (9%)       /testDir
4352B         4352B (100%)     /testFolder

free

The free command sends a request to the master to evict all blocks of a file from the Alluxio workers. If the argument to free is a directory, it will recursively free all files. This request is not guaranteed to take effect immediately, as readers may be currently using the blocks of the file. free will return immediately after the request is acknowledged by the master. Note that files must be already persisted in under storage before being freed or the free command will fail. Any pinned files cannot be freed unless -f option is specified. The free command does not delete any data from the under storage system, only removing the blocks of those files in Alluxio space to reclaim space. Metadata is not affected by this operation; a freed file will still show up if an ls command is run.

Usage: free [-f]

  • -f force to free files even pinned

For example, free can be used to manually manage Alluxio’s data caching.

$ ./bin/alluxio fs free /unused/data

getCapacityBytes

The getCapacityBytes command returns the maximum number of bytes Alluxio is configured to store.

For example, getCapacityBytes can be used to verify if your cluster is set up as expected.

$ ./bin/alluxio fs getCapacityBytes

getCmdStatus

The getCmdStatus command returns the detailed distributed command status based on a given JOB_CONTROL_ID. The detailed status includes:

  1. Successfully loaded or copied file paths.
  2. Statistics on the number of successful and failed file paths.
  3. Failed file paths, logged in a separate csv file.

For example, getCmdStatus can be used to check what files are loaded in a distributed command, and how many succeeded or failed.

$ ./bin/alluxio job getCmdStatus $JOB_CONTROL_ID
Sample Output:
Get command status information below:
Successfully loaded path $FILE_PATH_1
Successfully loaded path $FILE_PATH_2
Total completed file count is 2, failed file count is 0

getfacl

The getfacl command returns the ACL entries for a specified file or directory.

For example, getfacl can be used to verify that an ACL is changed successfully after a call to setfacl.

$ ./bin/alluxio fs getfacl /testdir/testfile

getSyncPathList

The getSyncPathList command gets all the paths that are under active syncing right now.

$ ./bin/alluxio fs getSyncPathList

getUsedBytes

The getUsedBytes command returns the number of used bytes in Alluxio.

For example, getUsedBytes can be used to monitor the health of the cluster.

$ ./bin/alluxio fs getUsedBytes

The head command prints the first 1 KB of data in a file to the console.

Using the -c [bytes] option will print the first n bytes of data to the console.

$ ./bin/alluxio fs head -c 2048 /output/part-00000

help

The help command prints the help message for a given fs subcommand. If the given command does not exist, it prints help messages for all supported subcommands.

Examples:

# Print all subcommands
$ ./bin/alluxio fs help

# Print help message for ls
$ ./bin/alluxio fs help ls

leader

The leader command prints the current Alluxio leading master hostname.

$ ./bin/alluxio fs leader

load

The load command moves data from the under storage system into Alluxio storage. If there is a Alluxio worker on the machine this command is run from, the data will be loaded to that worker. Otherwise, a random worker will be selected to serve the data.

If the data is already loaded into Alluxio, load is a no-op unless the --local flag is used. The --local flag forces the data to be loaded to a local worker even if the data is already available on a remote worker. If load is run on a directory, files in the directory will be recursively loaded.

For example, load can be used to prefetch data for analytics jobs.

$ ./bin/alluxio fs load /data/today

location

The location command returns the addresses of all the Alluxio workers which contain blocks belonging to the given file.

For example, location can be used to debug data locality when running jobs using a compute framework.

$ ./bin/alluxio fs location /data/2015/logs-1.txt

ls

The ls command lists all the immediate children in a directory and displays the file size, last modification time, and in memory status of the files. Using ls on a file will only display the information for that specific file.

The ls command will also load the metadata for any file or immediate children of a directory from the under storage system to Alluxio namespace if it does not exist in Alluxio. ls queries the under storage system for any file or directory matching the given path and creates a mirror of the file in Alluxio backed by that file. Only the metadata, such as the file name and size, are loaded this way and no data transfer occurs.

Options:

  • -d option lists the directories as plain files. For example, ls -d / shows the attributes of root directory.
  • -f option forces loading metadata for immediate children in a directory. By default, it loads metadata only at the first time at which a directory is listed. -f is equivalent to -Dalluxio.user.file.metadata.sync.interval=0.
  • -h option displays file sizes in human-readable formats.
  • -p option lists all pinned files.
  • -R option also recursively lists child directories, displaying the entire subtree starting from the input path.
  • --sort sorts the result by the given option. Possible values are size, creationTime, inMemoryPercentage, lastModificationTime, lastAccessTime and path.
  • -r reverses the sorting order.
  • --timestamp display the timestamp of the given option. Possible values are creationTime, lastModificationTime, and lastAccessTime. The default option is lastModificationTime.

For example, ls can be used to browse the file system.

$ ./bin/alluxio fs mount /s3/data s3://data-bucket/
# Loads metadata for all immediate children of /s3/data and lists them.
$ ./bin/alluxio fs ls /s3/data/

# Forces loading metadata.
$ aws s3 cp /tmp/somedata s3://data-bucket/somedata
$ ./bin/alluxio fs ls -f /s3/data

# Files are not removed from Alluxio if they are removed from the UFS (s3 here) only.
$ aws s3 rm s3://data-bucket/somedata
$ ./bin/alluxio fs ls -f /s3/data

Metadata sync is an expensive operation. A rough estimation is metadata sync on 1 million files will consume 2GB heap until the sync operation is complete. Therefore, we recommend not using forced sync to avoid accidental repeated sync operations. It is recommended to always specify a non-zero sync interval for metadata sync, so even if the sync is repeatedly triggered, the paths that have just been sync-ed can be identified and skipped.

# Should be avoided
$ ./bin/alluxio fs ls -f -R /s3/data

# Recommended. This will not sync files repeatedly in 1 minute.
$ ./bin/alluxio fs ls -Dalluxio.user.file.metadata.sync.interval=1min -R /s3/data

masterInfo

The masterInfo command prints information regarding master fault tolerance such as leader address, list of master addresses, and the configured Zookeeper address. If Alluxio is running in single master mode, masterInfo prints the master address. If Alluxio is running in fault tolerance mode, the leader address, list of master addresses and the configured Zookeeper address is printed.

For example, masterInfo can be used to print information regarding master fault tolerance.

$ ./bin/alluxio fs masterInfo

mkdir

The mkdir command creates a new directory in Alluxio space. It is recursive and will create any nonexistent parent directories. Note that the created directory will not be created in the under storage system until a file in the directory is persisted to the underlying storage. Using mkdir on an invalid or existing path will fail.

For example, mkdir can be used by an admin to set up the basic folder structures.

$ ./bin/alluxio fs mkdir /users
$ ./bin/alluxio fs mkdir /users/Alice
$ ./bin/alluxio fs mkdir /users/Bob

mount

The mount command links an under storage path to an Alluxio path, where files and folders created in Alluxio space under the path will be backed by a corresponding file or folder in the under storage path. For more details, see Unified Namespace.

Options:

  • --option <key>=<val> option passes an property to this mount point, such as S3 credentials
  • --readonly option sets the mount point to be readonly in Alluxio
  • --shared option sets the permission bits of the mount point to be accessible for all Alluxio users

Note that --readonly mounts are useful to prevent accidental write operations. If multiple Alluxio satellite clusters mount a remote storage cluster which serves as the central source of truth, --readonly option could help prevent any write operations on the satellite cluster from wiping out the remote storage.

For example, mount can be used to make data in another storage system available in Alluxio.

$ ./bin/alluxio fs mount /mnt/hdfs hdfs://host1:9000/data/
$ ./bin/alluxio fs mount --shared --readonly /mnt/hdfs2 hdfs://host2:9000/data/
$ ./bin/alluxio fs mount \
  --option s3a.accessKeyId=<accessKeyId> \
  --option s3a.secretKey=<secretKey> \
  /mnt/s3 s3://data-bucket/

mv

The mv command moves a file or directory to another path in Alluxio. The destination path must not exist or be a directory. If it is a directory, the file or directory will be placed as a child of the directory. mv is purely a metadata operation and does not affect the data blocks of the file. mv cannot be done between mount points of different under storage systems.

For example, mv can be used to re-organize your files.

$ ./bin/alluxio fs mv /data/2014 /data/archives/2014

persist

The persist command persists data in Alluxio storage into the under storage system. This is a server side data operation and will take time depending on how large the file is. After persist is complete, the file in Alluxio will be backed by the file in the under storage, and will still be available if the Alluxio blocks are evicted or otherwise lost.

Usage: persist [-p <parallelism>] [-t <timeout>] [-w <wait time>] <dir>

  • -p,--parallelism <parallelism>] (optional) Number of concurrent persist operations. (Default: 4)
  • -t,--timeout <timeout> (optional) Time in milliseconds for a single file persist to time out. (Default: 20 minutes)
  • -w,--wait <wait time> (optional) The time to wait in milliseconds before persisting. (Default: 0)
  • <dir> file or directory path in the Alluxio filesystem

If you are persisting multiple files, you can use the --parallelism <#> option to submit # of persist commands in parallel. For example, if your folder has 10,000 files, persisting with a parallelism factor of 10 will persist 10 files at a time until all 10,000 files are persisted.

For example, persist can be used after filtering a series of temporary files for the ones containing useful data.

$ ./bin/alluxio fs persist /tmp/experimental-logs-2.txt

pin

The pin command marks a file or folder as pinned in Alluxio. This is a metadata operation and will not cause any data to be loaded into Alluxio. If a file is pinned, any blocks belonging to the file will never be evicted from an Alluxio worker. If there are too many pinned files, Alluxio workers may run low on storage space, preventing other files from being cached.

For example, pin can be used to manually ensure performance if the administrator understands the workloads well.

$ ./bin/alluxio fs pin /data/today

policy

The policy command manages data migration policies set in Alluxio cluster. Policies enable Alluxio users to move data among under storage systems automatically based on various conditions.

To add a policy, use the following command:

./bin/alluxio fs policy add <PATH> "[<POLICY_NAME>:]ufsMigrate(<TIME_CONDITION>, <LOCATION>:<OPERATION>, ...)"

In the command above:

  • <POLICY_NAME> specifies the name of the policy, must be a unique string containing only alphanumeric characters. If the name is not specified, an automatically generated name will be used.
  • <PATH> specifies an Alluxio path where the policy will apply.
  • <TIME_CONDITION> specifies a time condition that the policy will be executed. Time condition is specified in the format of conditionType(startTime[, endTime]) There are three types of time condition:
    • olderThan: the policy will be executed when a file or directory is older than a certain period of time.
    • unusedFor: the policy will be executed when a file or directory is not used for a certain period of time.
    • unmodifiedFor: the policy will be executed when a file or directory is not modified for a certain period of time. The time period can be specified using a single start time, or a start time and an end time. For each time value, use s, m, h, d to indicate time unit seconds, minutes, hours, and days. For example,
      • unmodifiedFor(30m) sets a policy to execute after a file is not modified for 30 minutes.
      • olderThan(3h, 1d) sets a policy to execute after a file is older than 3 hours but younger than on day.
      • unusedFor(7d) sets a policy to execute after a file is not used for 7 days.
  • <LOCATION>:<OPERATION> specifies a file operation to be executed.
    • <LOCATION> specifies the target storage of the operation. Use UFS[<UFS_ALIAS>] to refer to a sub-UFS.
    • <OPERATION> specifies how the data is affected on this storage location. It can be either STORE to indicate the file data should be stored on this location, or REMOVE to indicate that data should be removed from this location.

To remove a policy, use the following command:

./bin/alluxio fs policy remove <POLICY_NAME>

where <POLICY_NAME> should be the name of the policy given in the add command. If a name is not given, you can check out the policy name using the list command.

To list all policies, use the following command:

./bin/alluxio fs policy list

To check the status of a policy, use the following command:

./bin/alluxio fs policy status <POLICY_NAME>

where <POLICY_NAME> should be the name of the policy that is listed in the policy list command or fsadmin report policy command.

To check the details of a specific action in the policy, locate the action id from either policy status command or fsadmin report policy command, and run:

bin/alluxio fs policy status <POLICY_NAME> --action <ACTION_ID>

Alternatively, the action details can also be fetched by the unionFS file path the action operated on:

bin/alluxio fs policy status <POLICY_NAME> --path <FILE_PATH>

You should only input one of these two options to specify the information you want otherwise actionId would be used to find the action information.

To manually trigger the scan on all the policies, use the following command.

/bin/alluxio fs policy trigger

For details on how to use policies, check out the Policy Driven Data Management page.

rm

The rm command removes a file from Alluxio space and the under storage system. The file will be unavailable immediately after this command returns, but the actual data may be deleted a while later.

  • Adding -R option deletes all contents of the directory and the directory itself.
  • Adding -U option skips the check for whether the UFS contents being deleted are in-sync with Alluxio before attempting to delete persisted directories. We recommend always using the -U option for the best performance and resource efficiency.
  • Adding --alluxioOnly option removes data and metadata from Alluxio space only. The under storage system will not be affected.
# Remove a file from Alluxio space and the under storage system
$ ./bin/alluxio fs rm /tmp/unused-file
# Remove a file from Alluxio space only
$ ./bin/alluxio fs rm --alluxioOnly /tmp/unused-file2

When deleting only from Alluxio but leaving the files in UFS, we recommend using -U and -Dalluxio.user.file.metadata.sync.interval=-1 to skip the metadata sync and the UFS check. This will save time and memory consumption on the Alluxio master.

$ bin/alluxio fs rm -R -U --alluxioOnly -Dalluxio.user.file.metadata.sync.interval=-1 /dir

When deleting a large directory (with millions of files) recursively both from Alluxio and UFS, the operation is expensive.

We recommend doing the deletion in the following way:

  1. Perform a direct sanity check against the UFS path with the corresponding file system API or CLI to make sure everything can be deleted safely. For example if the UFS is HDFS, use hdfs dfs -ls -R /dir to list the UFS files and check. We do not recommend doing this sanity check from Alluxio using a command like alluxio fs ls -R -f /dir, because the loaded file metadata will be deleted anyway, and the expensive metadata sync operation will essentially be wasted.

  2. Issue the deletion from Alluxio to delete files from both Alluxio and the UFS:

    # Disable the sync and skip the UFS check, to reduce memory consumption on the master side
$ bin/alluxio fs rm -R -U -Dalluxio.user.file.metadata.sync.interval=-1 /dir

Per 1 million files deleted, the memory overhead can be estimated as follows:

  • If both metadata sync and UFS check are disabled, recursively deleting from Alluxio only will hold 2GB JVM heap memory until the deletion completes.
  • If files are also deleted from UFS, there will not be extra heap consumption but the operation will take longer to complete.
  • If metadata sync is enabled, there will be another around 2GB overhead on the JVM heap until the operation completes.
  • If UFS check is enabled, there will another around 2GB overhead on the JVM heap until the operation completes.

Using this example as a guideline, estimate the total additional memory overhead as a proportion to the number of files to be deleted. Ensure that the leading master has sufficient available heap memory to perform the operation before issuing a large recursive delete command. A general good practice is to break deleting a large directory into deleting each individual children directories.

setfacl

The setfacl command modifies the access control list associated with a specified file or directory.

  • The-R option applies operations to all files and directories recursively.
  • The set option fully replaces the ACL while discarding existing entries. New ACL must be a comma separated list of entries, and must include user, group, and other for compatibility with permission bits.
  • The -m option modifies the ACL by adding/overwriting new entries.
  • The -x option removes specific ACL entries.
  • The -b option removes all ACL entries, except for the base entries.
  • The -d option indicates that operations apply to the default ACL
  • The -k option removes all the default ACL entries.

For example, setfacl can be used to give read and execute permissions to a user named testuser.

$ ./bin/alluxio fs setfacl -m "user:testuser:r-x" /testdir/testfile

setReplication

The setReplication command sets the max and/or min replication level of a file or all files under a directory recursively. This is a metadata operation and will not cause any replication to be created or removed immediately. The replication level of the target file or directory will be changed automatically in background.

  • The --min option specifies the minimal replication level
  • The --max optional specifies the maximal replication level. Specify -1 as the argument of --max option to indicate no limit of the maximum number of replicas.
  • If the specified path is a directory and -R is specified, it will recursively set all files in this directory.

For example, setReplication can be used to ensure the replication level of a file has at least one copy and at most three copies in Alluxio:

$ ./bin/alluxio fs setReplication --max 3 --min 1 /foo

setTtl

The setTtl command sets the time-to-live of a file or a directory, in milliseconds. If set TTL is run on a directory, the same TTL attributes is set on all its children. If a directory’s TTL expires, all its children will also expire.

Action parameter --action will indicate the action to perform once the file or directory expires. The default action, delete, deletes the file or directory from both Alluxio and the under storage system, whereas the action free frees the file from Alluxio even if pinned.

For example, setTtl with action delete cleans up files the administrator knows are unnecessary after a period of time, or with action free just remove the contents from Alluxio to make room for more space in Alluxio.

# After 1 day, delete the file in Alluxio and UFS
$ ./bin/alluxio fs setTtl /data/good-for-one-day 86400000
# After 1 day, free the file from Alluxio
$ ./bin/alluxio fs setTtl --action free /data/good-for-one-day 86400000

startSync

The startSync command starts the automatic syncing process of the specified path.

$ ./bin/alluxio fs startSync /data/2014

stat

The stat command dumps the FileInfo representation of a file or a directory to the console. It is primarily intended to assist power users in debugging their system. Generally viewing the file info in the UI will be easier to understand.

One can specify -f <arg> to display info in given format:

  • %N: name of the file
  • %z: size of file in bytes
  • %u: owner
  • %g: group name of owner
  • %y or %Y: modification time, where %y shows the UTC date in the form yyyy-MM-dd HH:mm:ss and %Y shows the number of milliseconds since January 1, 1970 UTC
  • %b: Number of blocks allocated for file

For example, stat can be used to debug the block locations of a file. This is useful when trying to achieve locality for compute workloads.

# Displays file's stat
$ ./bin/alluxio fs stat /data/2015/logs-1.txt

# Displays directory's stat
$ ./bin/alluxio fs stat /data/2015

# Displays the size of file
$ ./bin/alluxio fs stat -f %z /data/2015/logs-1.txt

stopSync

The stopSync command stops the automatic syncing process of the specified path.

$ ./bin/alluxio fs stopSync /data/2014

tail

The tail command outputs the last 1 KB of data in a file to the console. Using the -c [bytes] option will print the last n bytes of data to the console.

For example, tail can be used to verify the output of a job is in the expected format or contains expected values.

$ ./bin/alluxio fs tail /output/part-00000

test

The test command tests a property of a path, returning 0 if the property is true or 1 otherwise.

Options:

  • -d option tests whether path is a directory.
  • -e option tests whether path exists.
  • -f option tests whether path is a file.
  • -s option tests whether path is not empty.
  • -z option tests whether file is zero length.

Examples:

$ ./bin/alluxio fs test -d /someDir
$ echo $?

touch

The touch command creates a 0-byte file. Files created with touch cannot be overwritten and are mostly useful as flags.

For example, touch can be used to create a file signifying the completion of analysis on a directory.

$ ./bin/alluxio fs touch /data/yesterday/_DONE_

unmount

The unmount command disassociates an Alluxio path with an under storage directory. Alluxio metadata for the mount point is removed along with any data blocks, but the under storage system will retain all metadata and data. See Unified Namespace for more details.

For example, unmount can be used to remove an under storage system when the users no longer need data from that system.

$ ./bin/alluxio fs unmount /s3/data

If there are files under the mount point, the unmount operation will implicitly delete those files from Alluxio. See the rm command for how to estimate the memory consumption. It is recommended to remove those files in Alluxio first, before the unmount.

unpin

The unpin command unmarks a file or directory in Alluxio as pinned. This is a metadata operation and will not evict or delete any data blocks. Once a file is unpinned, its data blocks can be evicted from the various Alluxio workers containing the block.

For example, unpin can be used when the administrator knows there is a change in the data access pattern.

$ ./bin/alluxio fs unpin /data/yesterday/join-table

unsetTtl

The unsetTtl command will remove the TTL attributes of a file or directory in Alluxio. This is a metadata operation and will not evict or store blocks in Alluxio. The TTL of a file can later be reset with setTtl.

For example, unsetTtl can be used if a regularly managed file requires manual management.

$ ./bin/alluxio fs unsetTtl /data/yesterday/data-not-yet-analyzed

updateMount

The updateMount command updates options for a mount point while keeping the Alluxio metadata under the path.

Usage: updateMount [--readonly] [--shared] [--option <key=val>] <alluxioPath>

  • --readonly (optional) mount point is readonly in Alluxio
  • --shared (optional) mount point is shared
  • --option <key>=<val> (optional) options for this mount point. For security reasons, no options from existing mount point will be inherited.
  • <alluxioPath> Directory path in the Alluxio filesystem

Table Operations

$ ./bin/alluxio table
Usage: alluxio table [generic options]
	 [attachdb [-o|--option <key=value>] [--db <alluxio db name>] [--ignore-sync-errors] <udb type> <udb connection uri> <udb db name>]
	 [detachdb <db name>]
	 [ls [<db name> [<table name>]]]
	 [sync <db name>]
	 [transform <db name> <table name>]
	 [transformStatus [<job ID>]]

The table subcommand manages the structured data service of Alluxio.

Note: This command requires the Alluxio cluster to be running.

attachdb

Syntax:

attachdb [-o|--option <key=value>] [--db <alluxio db name>] [--ignore-sync-errors] <udb type> <udb connection uri> <udb db name>

The attachdb command attaches an existing “under database” to the Alluxio catalog. This is analogous to mounting a under filesystem to the Alluxio filesystem namespace. Once a database is attached, it will be exposed through the Alluxio catalog. Here is an example of the usage:

$ ./bin/alluxio table attachdb hive thrift://HOSTNAME:9083 hive_db_name

This command will attach the database hive_db_name (of type hive) from the URI thrift://HOSTNAME:9083 to the Alluxio catalog, using the same database name hive_db_name.

Here are the attach command options:

  • --db <alluxio db name>: specify a different Alluxio database name
  • --ignore-sync-errors: ignore sync errors, and keeps the database attached
  • -o|--option <key=value>: (multiple) additional properties associated with the attached db and UDB

Here are the additional properties possible for the -o options:

  • udb-<UDB_TYPE>.mount.option.{<UFS_PREFIX>}.<MOUNT_PROPERTY>: specify a mount option for a particular UFS path
    • <UDB_TYPE>: the UDB type
    • <UFS_PREFIX>: the UFS path prefix, or a regex string starts with regex: that the mount properties are for
    • <MOUNT_PROPERTY>: an Alluxio mount property
  • catalog.db.config.file: the config file for the UDB, you can configure which tables and partitions to bypass from Alluxio in a configuration specified by this option. See UDB Configuration File for details.
  • catalog.db.ignore.udb.tables: comma-separated list of table names to ignore from the UDB
  • catalog.db.sync.threads: number of parallel threads to use to sync with the UDB. If too large, the sync may overload the UDB, and if set too low, syncing a database with many tables make take a long time. The default is 4.

Hive UDB

For the hive udb type, during the attach process, the Alluxio catalog will auto-mount all the table/partition locations in the specified database, to Alluxio. You can supply the mount options for the possible table locations with the option -o udb-hive.mount.option.{scheme/authority}.key=value or -o udb-hive.mount.option.{regex:REGEX}.key=value

$ ./bin/alluxio table attachdb hive thrift://HOSTNAME:9083 hive_db_name --db=alluxio_db_name  \
  -o udb-hive.mount.option.{s3a://bucket1}.s3a.accessKeyId=abc \
  -o udb-hive.mount.option.{s3a://bucket2}.s3a.accessKeyId=123

This command will attach the database hive_db_name (of type hive) from the URI thrift://HOSTNAME:9083 to the Alluxio catalog, using the same database name alluxio_db_name. When paths are mounted for s3a://bucket1, the mount option s3a.accessKeyId=abc will be used, and when paths are mounted for s3a://bucket2, the mount option s3a.accessKeyId=123 will be used.

Or using regex expression if the options are same the two buckets.

$ ./bin/alluxio table attachdb hive thrift://HOSTNAME:9083 hive_db_name --db=alluxio_db_name  \
  -o udb-hive.mount.option.{regex:s3a://bucket.*}.s3a.accessKeyId=abc

Besides mount options, there are some additional properties with the -o options:

  • udb-hive.<UDB_PROPERTY>: specify the UDB options for the Hive UDB. The options are as follows
    • allow.diff.partition.location.prefix: Whether to mount partitions that do not share the same location prefix with table location(true/false, default false)

Glue UDB

For glue udb type, there are some additional properties with the -o options:

  • udb-glue.<UDB_PROPERTY>: specify the UDB options for the Glue UDB. The options are as follows:
    • aws.region: the glue aws region
    • aws.catalog.id: the aws catalog id
    • aws.accesskey: the aws access key id
    • aws.secretkey: the aws secret key
    • aws.proxy.protocol: The protocol(HTTP/HTTPS) to use for connecting to the proxy server
    • aws.proxy.host: The proxy host the client will connect through
    • aws.proxy.port: The proxy port the client will connect through
    • aws.proxy.username: The proxy user name
    • aws.proxy.password: The proxy password
    • table.column.statistics: Enable table column statistics(true/false)
    • partition.column.statistics: Enable partition column statistics(true/false)

You can supply the mount options for the glue as follows:

$ ./bin/alluxio table attachdb --db alluxio_db_name glue null glue_db_name \
    -o udb-glue.aws.region=<AWS_GLUE_REGION> \
    -o udb-glue.aws.catalog.id=<AWS_CATALOGID> \
    -o udb-glue.aws.accesskey=<AWS_ACCESSKEY_ID> \
    -o udb-glue.aws.secretkey=<AWS_SERCRETKEY_ID>

This command will attach the database glue_db_name (of type glue) to the Alluxio catalog, using the same database name alluxio_db_name. Please notice that glue udb does not need the URI as hive udb. When glue udb access to AWS glue, the aws region udb-glue.aws.region, AWS catalog id udb-glue.aws.catalog.id and AWS credentials, udb-glue.aws.accesskey and udb-glue.aws.secretkey , need to be provided.

detachdb

The detachdb command is the opposite of the attachdb command. Detaching a database will remove the connection to the under database, and remove it from the Alluxio catalog. Example usage:

$ ./bin/alluxio table detachdb alluxio_db_name

This command will detach the database name alluxio_db_name from the Alluxio catalog.

ls

The ls command shows information about the Alluxio catalog. Here are some examples:

$ ./bin/alluxio table ls

This command without any arguments will show all the databases attached in the system.

$ ./bin/alluxio table ls db_name

This command with 1 argument will show all the tables in the db_name database.

$ ./bin/alluxio table ls db_name table_name

This command with 2 arguments will show the table information of the table_name table in the db_name database.

sync

The sync command syncs the metadata of specified database name with the under database. Here is an example:

$ ./bin/alluxio table sync db_name

This will sync the metadata of db_name database name with its under database. The sync will update, add, remove catalog metadata according to the changes found in the underlying database and tables. For example, if the under database is hive, and the metadata of its tables is updated in the Hive Metastore (like MSCK REPAIR or other commands), then this sync command will update the Alluxio metadata with the updated Hive metadata. If an existing Alluxio partition or table is updated and previously had a transformation, then the transformation is invalidated, and must be re-triggered via the transform command.

If the metadata is NOT updated in the under database, then this sync command will not update the Alluxio catalog metadata, even if the data of the table has been updated. For example, if files are added to a Hive table but the Hive Metastore is not updated, the sync will not detect changes to the metadata.

transform

The transform command will transform a table for improved efficiency when reading the table. Here is an example usage:

$ ./bin/alluxio table transform db_name table_name [-d <definition>]

This command will invoke a transformation on the table. The transformation is performed asynchronously, and will coalesce to a fewer number of files, and convert into the parquet file format.

In 2.1.0, the supported file formats which can be transformed are: parquet and csv file formats. The resulting transformations are in the parquet file format. Additional formats for input and output will be implemented in future versions. For the coalesce feature, by default it will coalesce into a maximum of 100 files, with each file no smaller than 2GB.

The definition format takes a form of configuration separated by semicolon and specifies the details of the output format. Available configurations are:

file.count.max=<num> (maximum number of files in transformed output)
file.size.min=<num> (minimum file size in bytes of the output)
sort=<columns> (comma separated list of columns)

transformStatus

The transformStatus command will display information about a table transformation. Here is an example usage:

$ ./bin/alluxio table transformStatus transform_id

This command will display status details on the transformation identified by transform_id.