cbft command-line
Command-line help is available by running...
./cbft --help
That will print out usage/help output...
cbft: couchbase full-text server
Usage: cbft [flags]
Flags:
-b, -bindHttp ADDR:PORT
local address:port where this node will listen and
serve HTTP/REST API requests and the web-based
admin UI; default is '0.0.0.0:8095'.
-c, -cfg, -cfgConnect CFG_CONNECT
connection string to a configuration provider/server
for clustering multiple cbft nodes:
* couchbase:http://BUCKET_USER:BUCKET_PSWD@CB_HOST:CB_PORT
- manages a cbft cluster configuration in a couchbase
3.x bucket; for example:
'couchbase:http://my-cfg-bucket@127.0.0.1:8091';
* simple
- intended for development usage, the 'simple'
configuration provider manages a configuration
for a single, unclustered cbft node in a local
file that's stored in the dataDir;
* metakv
- manages a cbft cluster configuration in couchbase metakv store;
environment variable CBAUTH_REVRPC_URL needs to be set
for metakv; for example:
'export CBAUTH_REVRPC_URL=http://user:password@localhost:9000/cbft';
default is 'simple'.
-container PATH
optional slash separated path of logical parent containers
for this node, for shelf/rack/row/zone awareness.
-data, -dataDir DIR
optional directory path where local index data and
local config files will be stored for this node;
default is 'data'.
-e, -extra extra info
extra info you want stored with this node
-h, -H, -?, -help
print this usage message and exit.
-register STATE
optional flag to register this node in the cluster as:
* wanted - make node wanted in the cluster,
if not already, so that it will participate
fully in data operations;
* wantedForce - same as wanted, but forces a cfg update;
* known - make node known to the cluster,
if not already, so it will be admin'able
but won't yet participate in data operations;
this is useful for staging several nodes into
the cluster before making them fully wanted;
* knownForce - same as known, but forces a cfg update;
* unwanted - make node unwanted, but still known to the cluster;
* unknown - make node unwanted and unknown to the cluster;
* unchanged - don't change the node's registration state;
default is 'wanted'.
-s, -server URL
URL to datasource server; example when using couchbase 3.x as
your datasource server: 'http://localhost:8091';
use '.' when there is no datasource server.
-staticDir DIR
optional directory for web UI static content;
default is using the static resources embedded
in the program binary.
-staticETag ETAG
optional ETag for web UI static content.
-tags TAGS
optional comma-separated list of tags or enabled roles
for this node, such as:
* feed - node can connect feeds to datasources;
* janitor - node can run a local janitor;
* pindex - node can maintain local index partitions;
* planner - node can replan cluster-wide resource allocations;
* queryer - node can execute queries;
default is ("") which means all roles are enabled.
-uuid UUID
optional uuid for this node; by default, a previous uuid file
is read from the dataDir, or a new uuid is auto-generated
and saved into the dataDir.
-v, -version
print version string and exit.
-weight INTEGER
optional weight of this node, where a more capable
node should have higher weight; default is 1.
Examples:
Getting started, using a couchbase (3.x) on localhost as the datasource:
./cbft -server=http://localhost:8091
Example where cbft's configuration is kept in a couchbase "cfg-bucket":
./cbft -cfg=couchbase:http://cfg-bucket@CB_HOST:8091 \
-server=http://CB_HOST:8091
See also: http://github.com/couchbaselabs/cbft
Copyright (c) 2015 Couchbase, Inc.