Skip to content

Latest commit

 

History

History
275 lines (212 loc) · 8.46 KB

moray.md

File metadata and controls

275 lines (212 loc) · 8.46 KB
Raw

moray 1 "January 2017" Moray "Moray Client Tools"

NAME

moray - command-line tools for Moray

DESCRIPTION

Moray is a JSON-based key-value store. Users can create buckets, each containing any number of objects indexed by a primary key. Additional indexes can be specified with each bucket. Moray servers are typically stateless components deployed atop PostgreSQL databases, where buckets correspond to database tables, objects correspond to rows, and database indexes are created for each bucket index.

The moray npm package contains a set of command-line tools to allow users to create, update, and delete both buckets and objects. These tools connect to one or more remote Moray servers over TCP and execute operations.

Working with buckets:

  • putbucket: create or update a bucket
  • getbucket: print detailed information about one bucket
  • listbuckets: print detailed information about all buckets
  • delbucket: delete a bucket and all of its contents

Working with objects:

  • putobject: create or update an object
  • getobject: fetch the contents of an object by primary key
  • delobject: delete an object by primary key
  • findobjects: fetch the contents of multiple objects using a filter
  • delmany: delete multiple objects using a filter
  • updatemany: update multiple objects using a filter
  • reindexobjects: populate a newly-added index

Working with remote servers:

  • morayping: check whether Moray is online
  • morayversion: check the version of a Moray server
  • sql: execute a SQL string on a Moray server
  • gettokens: fetch the list of shards from electric-moray

The tools in this package support two modes for locating the remote Moray server on which to operate:

  • Using the -S/--service SERVICE_NAME option or the MORAY_SERVICE environment variable, users specify a DNS domain to which SRV records are attached that describe the list of instances available. SRV records provide both a name for the host (which may be an IP address or another DNS domain) and a port on which to connect over TCP. This mode is preferred for general use because it provides information about all instances and allows the client to balance multiple requests across different, equivalent servers.
  • Using the -h/--host HOST_OR_IP and -p/--port PORT options or the MORAY_URL environment variable, users specify a specific IP address or DNS domain to which traditional name records are attached and a TCP port to which to connect. This is useful primarily for testing against specific server instances.

If the -S/--service SERVICE_NAME command-line option is specified, it is always used directly as described above.

If the -h/--host HOST_OR_IP or -p/--port PORT options are specified, they are used directly as described above. If one is specified and not the other, then the other value is filled in from the MORAY_URL environment variable. Otherwise, defaults of IP 127.0.0.1 and port 2020 are used.

If none of these command-line options are specified:

  • if MORAY_SERVICE is specified, it is used to invoke the first mode
  • if MORAY_URL is specified, is used to invoke the second mode
  • if neither is specified, the second mode is invoked with default values 127.0.0.1 port 2020.

OPTIONS

The following COMMON_OPTIONS options are accepted by all of these commands:

-b, --bootstrap-domain BOOTSTRAP_DOMAIN Specifies the domain name for the nameservers themselves. Triton and Manta both provide domain names for the nameservers themselves. This is useful in split DNS environments to ensure that the Moray only uses the nameservers that know about the target service. This applies to both SRV-record-based discovery and traditional A-record-based discovery, but has no impact when connecting to a specific IP address and port.

-h, --host HOST_OR_IP Specifies an IP address or DNS domain for the remote Moray server. See above for details.

-p, --port PORT Specifies the TCP port for the remote Moray server. See above for details.

-S, --service SERVICE Specifies a DNS domain to be used for SRV-based service discovery of the remote Moray server. See above for details. SERVICE must not be an IP address.

-v, --verbose Increases the verbosity of the built-in bunyan logger. By default, the logger is created with bunyan level fatal. Each additional use of -v increases the verbosity by one level (to error, warn, and so on). Log messages are emitted to stderr. See also the LOG_LEVEL environment variable.

ENVIRONMENT

LOG_LEVEL Sets the node-bunyan logging level. Defaults to "fatal".

MORAY_BOOTSTRAP_DOMAIN Used as a fallback value for the -b/--bootstrap-domain option.

MORAY_SERVICE Used as a fallback value for -S/--service if neither of -h/--host or -p/--port is specified.

MORAY_URL A URL of the form tcp://HOSTNAME_OR_IP[:PORT] where the specified HOSTNAME_OR_IP and PORT will be used as fallback values for the -h/--host or -p/--port options, respectively. This value is only used if MORAY_SERVICE is not present in the environment and at least one of the -h/--host or -p/--port options is not specified.

EXIT STATUS

0 Indicates successful completion

1 Indicates failure

2 Indicates an invalid invocation (usage error)

EXAMPLES

Create a bucket for "accounts" version 1 with unique "loginName" and unique numeric values for "uid":

$ putbucket -x 1 -u loginName -u uid:number accounts

Insert a few accounts. The keys will match the login names:

$ putobject -d '{ "loginName": "lisa", "uid": "800", "country": "USA" }' \
    accounts lisa
$ putobject -d '{ "loginName": "hugh", "uid": "801", "country": "UK" }' \
    accounts hugh

Fetch one of these back:

$ getobject accounts lisa
{
  "bucket": "accounts",
  "key": "lisa",
  "value": {
    "loginName": "lisa",
    "uid": 800,
    "country": "USA"
  },
  "_id": 1,
  "_etag": "77472568",
  "_mtime": 1485539314987,
  "_txn_snap": null,
  "_count": null
}

Fetch objects having uids at most 800:

$ findobjects accounts '(uid<=800)'
{
  "bucket": "accounts",
  "key": "lisa",
  "value": {
    "loginName": "lisa",
    "uid": 800
  },
  "_id": 1,
  "_etag": "77472568",
  "_mtime": 1485539314987,
  "_txn_snap": null,
  "_count": 1
}

Update the bucket to version 2, which contains a new non-unique indexed column for "country":

$ putbucket -x 2 -u loginName -u uid:number -i country accounts

Re-index the objects in the bucket:

$ reindexobjects accounts
bucket "accounts": 2 objects processed (continuing)
bucket "accounts": all objects processed

Now we can search for accounts by country:

$ findobjects accounts '(country=UK)'
{
  "bucket": "accounts",
  "key": "hugh",
  "value": {
    "loginName": "hugh",
    "uid": 801,
    "country": "UK"
  },
  "_id": 2,
  "_etag": "82E66E74",
  "_mtime": 1485539319781,
  "_txn_snap": null,
  "_count": 1
}

Update "country" for accounts with country = "UK":

$ updatemany -d '{ "country": "United Kingdom" }' accounts '(country=UK)'
{ count: 1, etag: 'ue6d321d' }

Now fetch back "hugh":

$ getobject accounts hugh
{
  "bucket": "accounts",
  "key": "hugh",
  "value": {
    "loginName": "hugh",
    "uid": 801,
    "country": "United Kingdom"
  },
  "_id": 2,
  "_etag": "ue6d321d",
  "_mtime": 1485539410157,
  "_txn_snap": null,
  "_count": null
}

Now delete that object:

$ delobject accounts hugh

List everything in the bucket (more precisely, everything having a non-null "loginName"):

$ findobjects accounts '(loginName=*)'
{
  "bucket": "accounts",
  "key": "lisa",
  "value": {
    "loginName": "lisa",
    "uid": 800,
    "country": "USA"
  },
  "_id": 1,
  "_etag": "77472568",
  "_mtime": 1485539314987,
  "_txn_snap": null,
  "_count": 1
}

Finally, remove the bucket:

$ delbucket accounts

SEE ALSO

moray(3), delbucket(1), delmany(1), delobject(1), findobjects(1), getbucket(1), getobject(1), gettokens(1), listbuckets(1), morayping(1), morayversion(1), putbucket(1), putobject(1), reindexobjects(1), sql(1), updatemany(1), jsontool

DIAGNOSTICS

See the -v/--verbose option and the LOG_LEVEL environment variable.