commands
Command implementation.
Provides ASCII-based table with minimal styling support.
Command line Visor.
Singleton companion object.
Static context.
Visor 'ack' command implementation.
Visor 'ack' command implementation.
+-------------------------------------------+ | ack | Acks arguments on all remote nodes. | +-------------------------------------------+
ack {"s"}
ack ("s", f)s
Optional string to print on each remote node.
f
Optional Scala predicate on 'ScalarRichNodePimp' filtering nodes in the topology.ack "Howdy!" Prints 'Howdy!' on all nodes in the topology. ack("Howdy!", _.id8.startsWith("123")) Prints 'Howdy!' on all nodes satisfying this predicate. ack Prints local node ID on all nodes in the topology.
Visor 'alert' command implementation.
Visor 'alert' command implementation.
+---------------------------------------------------------------------+
| alert | Generates email alerts for user-defined events. |
| | Node events and grid-wide events are defined via mnemonics. |
+---------------------------------------------------------------------+alert alert "-u {-id=|-a}" alert "-r {-t=} -c1=e1 -c2=e2 ... -ck=ek "
-u
Unregisters alert(s). Either '-a' flag or '-id' parameter is required.
Note that only one of the '-u' or '-r' is allowed.
If neither '-u' or '-r' provided - all alerts will be printed.
-a
When provided with '-u' - all alerts will be unregistered.
-id=<alert-id>
When provided with '-u' - alert with matching ID will be unregistered.
-r
Register new alert with mnemonic predicate(s).
Note that only one of the '-u' or '-r' is allowed.
If neither '-u' or '-r' provided - all alerts will be printed.
-t
Defines notification frequency in seconds. Default is 15 minutes.
This parameter can only appear with '-r'.
-ck=ek<num>
This defines a mnemonic for the metric that will be measured:
Grid-wide metrics (not node specific):
-cc Total number of available CPUs in the grid.
-nc Total number of nodes in the grid.
-hc Total number of physical hosts in the grid.
-cl Current average CPU load (in %) in the grid.
Per-node current metrics:
-aj Active jobs on the node.
-cj Cancelled jobs on the node.
-tc Thread count on the node.
-it Idle time on the node.
Note: <num> can have 's', 'm', or 'h' suffix indicating
seconds, minutes, and hours. By default (no suffix provided)
value is assumed to be in milliseconds.
-ut Up time on the node.
Note: <num> can have 's', 'm', or 'h' suffix indicating
seconds, minutes, and hours. By default (no suffix provided)
value is assumed to be in milliseconds.
-je Job execute time on the node.
-jw Job wait time on the node.
-wj Waiting jobs count on the node.
-rj Rejected jobs count on the node.
-hu Heap memory used (in MB) on the node.
-cd Current CPU load on the node.
-hm Heap memory maximum (in MB) on the node.
Comparison part of the mnemonic predicate:
=eq<num> Equal '=' to '' number.
=neq<num> Not equal '!=' to '' number.
=gt<num> Greater than '>' to '' number.
=gte<num> Greater than or equal '>=' to '' number.
=lt<num> Less than '<' to 'NN' number.
=lte<num> Less than or equal '<=' to '' number.
NOTE: Email notification will be sent for the alert only when all
provided mnemonic predicates evaluate to 'true'.alert
Prints all currently registered alerts.
alert "-u -a"
Unregisters all currently registered alerts.
alert "-u -id=12345678"
Unregisters alert with provided ID.
alert "-r -t=900 -cc=gte4 -cl=gt50"
Notify every 15 min if grid has >= 4 CPUs and > 50% CPU load.
Visor 'cache' command implementation.
Visor 'cache' command implementation.
+-----------------------------------------------------------------------------------------+ | cache | Prints statistics about caches from specified node on the entire grid. | | | Output sorting can be specified in arguments. | | | | | | Output abbreviations: | | | # Number of nodes. | | | H/h Number of cache hits. | | | M/m Number of cache misses. | | | R/r Number of cache reads. | | | W/w Number of cache writes. | +-----------------------------------------------------------------------------------------+ | cache -compact | Compacts all entries in cache on all nodes. | +-----------------------------------------------------------------------------------------+ | cache -clear | Clears all entries from cache on all nodes. | +-----------------------------------------------------------------------------------------+ | cache -scan | List all entries in cache with specified name. | +-----------------------------------------------------------------------------------------+
cache
cache -i
cache {-c=<cache-name>} {-id=<node-id>|id8=<node-id8>} {-s=lr|lw|hi|mi|re|wr} {-a} {-r}
cache -clear {-c=<cache-name>}
cache -compact {-c=<cache-name>}
cache -scan -c=<cache-name> {-id=<node-id>|id8=<node-id8>} {-p=<page size>}-id=<node-id>
Full ID of the node to get cache statistics from.
Either '-id8' or '-id' can be specified.
If neither is specified statistics will be gathered from all nodes.
-id8=<node-id>
ID8 of the node to get cache statistics from.
Either '-id8' or '-id' can be specified.
If neither is specified statistics will be gathered from all nodes.
-c=<cache-name>
Name of the cache.
-s=lr|lw|hi|mi|re|wr|cn
Defines sorting type. Sorted by:
lr Last read.
lw Last write.
hi Hits.
mi Misses.
rd Reads.
wr Writes.
If not specified - default sorting is 'lr'.
-i
Interactive mode.
User can interactively select node for cache statistics.
-r
Defines if sorting should be reversed.
Can be specified only with '-s' argument.
-a
Prints details statistics about each cache.
By default only aggregated summary is printed.
-compact
Compacts entries in cache.
-clear
Clears cache.
-scan
Prints list of all entries from cache.
-p=<page size>
Number of object to fetch from cache at once.
Valid range from 1 to 100.
By default page size is 25.cache
Prints summary statistics about all caches.
cache -id8=12345678 -s=hi -r
Prints summary statistics about caches from node with specified id8
sorted by number of hits in reverse order.
cache -i
Prints cache statistics for interactively selected node.
cache -s=hi -r -a
Prints detailed statistics about all caches sorted by number of hits in reverse order.
cache -compact
Compacts entries in interactively selected cache.
cache -compact -c=cache
Compacts entries in cache with name 'cache'.
cache -clear
Clears interactively selected cache.
cache -clear -c=cache
Clears cache with name 'cache'.
cache -scan
Prints list entries from interactively selected cache.
cache -scan -c=cache
Prints list entries from cache with name 'cache' from all nodes with this cache.
cache -scan -c=@c0 -p=50
Prints list entries from cache with name taken from 'c0' memory variable
with page of 50 items from all nodes with this cache.
cache -scan -c=cache -id8=12345678
Prints list entries from cache with name 'cache' and node '12345678' ID8.
Visor 'config' command implementation.
Visor 'config' command implementation.
+-------------------------------------+ | config | Prints node configuration. | +-------------------------------------+
config
config "{-id=|id8=}" -id=<node-id>
Full node ID. Either '-id8' or '-id' can be specified.
If neither is specified - command starts in interactive mode.
-id8=<node-id8>
Node ID8. Either '-id8' or '-id' can be specified.
If neither is specified - command starts in interactive mode.config "-id8=12345678" Prints configuration for node with '12345678' ID8. config Starts command in interactive mode.
Visor 'deploy' command implementation.
Visor 'deploy' command implementation.
+---------------------------------------------------+ | deploy | Copies file or directory to remote host. | | | Command relies on SFTP protocol. | +---------------------------------------------------+
deploy "-h={{:}@}{:} {-u=}
{-p=} {-k=} -s= {-d}" -h={<username>{:<password>}@}<host>{:<port>}
Host specification.
<host> can be a hostname, IP or range of IPs.
Example of range is 192.168.1.100~150,
which means all IPs from 192.168.1.100 to 192.168.1.150 inclusively.
Default port number is 22.
This option can be provided multiple times.
-u=<username>
Default username.
Used if specification doesn't contain username.
If default is not provided as well, current local username will be used.
-p=<password>
Default password.
Used if specification doesn't contain password.
If default is not provided as well, it will be asked interactively.
-k=<path>
Path to private key file.
If provided, it will be used for all specifications that doesn't contain password.
-s=<path>
Source path.
-d=<path>
Destination path (relative to GRIDGAIN_HOME).
If not provided, files will be copied to the root of GRIDGAIN_HOME.deploy "-h=uname:passwd@host -s=/local/path -d=/remote/path" Copies file or directory to remote host (password authentication). deploy "-h=uname@host -k=ssh-key.pem -s=/local/path -d=/remote/path" Copies file or directory to remote host (private key authentication).
Visor 'disco' command implementation.
Visor 'disco' command implementation.
+---------------------------------------------------------------------------------------+ | disco | Prints topology change log as seen from the oldest node. | | | Timeframe for quering events can be specified in arguments. | | | | | | Note that this command depends on GridGain events. | | | | | | GridGain events can be individually enabled and disabled and disabled events | | | can affect the results produced by this command. Note also that configuration | | | of Event Storage SPI that is responsible for temporary storage of generated | | | events on each node can also affect the functionality of this command. | | | | | | By default - all events are enabled and GridGain stores last 10,000 local | | | events on each node. Both of these defaults can be changed in configuration. | +---------------------------------------------------------------------------------------+
disco
disco "{-t=s|m|h|d} {-r} {-c=}" -t=<num>s|m|h|d
Defines timeframe for querying events:
=<num>s Events fired during last <num> seconds.
=<num>m Events fired during last <num> minutes.
=<num>h Events fired during last <num> hours.
=<num>d Events fired during last <num> days.
-r
Defines whether sorting should be reversed.
-c=<n>
Defines the maximum events count that can be shown.disco
Prints all discovery events sorted chronologically (oldest first).
disco "-r"
Prints all discovery events sorted chronologically in reversed order (newest first).
disco "-t=2m"
Prints discovery events fired during last two minutes sorted chronologically.
Visor 'events' commands implementation.
Visor 'events' commands implementation.
+----------------------------------------------------------------------------------------+ | events | Print events from a node. | | | | | | Note that this command depends on GridGain events. | | | | | | GridGain events can be individually enabled and disabled and disabled events | | | can affect the results produced by this command. Note also that configuration | | | of Event Storage SPI that is responsible for temporary storage of generated | | | events on each node can also affect the functionality of this command. | | | | | | By default - all events are enabled and GridGain stores last 10,000 local | | | events on each node. Both of these defaults can be changed in configuration. | +----------------------------------------------------------------------------------------+
events
events "{-id=|-id8=} {-e=}
{-t=s|m|h|d} {-s=e|t} {-r} {-c=}" -id=<node-id>
Full node ID.
Either '-id' or '-id8' can be specified.
If called without the arguments - starts in interactive mode.
-id8
Node ID8.
Either '-id' or '-id8' can be specified.
If called without the arguments - starts in interactive mode.
-e=<ch,de,di,jo,ta,cl,ca,sw>
Comma separated list of event types that should be queried:
ch Checkpoint events.
de Deployment events.
di Discovery events.
jo Job execution events.
ta Task execution events.
ca Cache events.
cp Cache pre-loader events.
sw Swapspace events.
-t=<num>s|m|h|d
Defines time frame for querying events:
=<num>s Queries events fired during last <num> seconds.
=<num>m Queries events fired during last <num> minutes.
=<num>h Queries events fired during last <num> hours.
=<num>d Queries events fired during last <num> days.
-s=e|t
Defines sorting of queried events:
=e Sorted by event type.
=t Sorted chronologically.
Only one '=e' or '=t' can be specified.
-r
Defines if sorting should be reversed.
Can be specified only with -s argument.
-c=<n>
Defines the maximum events count that can be shown.
Values in summary tables are calculated over the whole list of events.events "-id8=12345678" Queries all events from node with '12345678' ID8. events "-id8=12345678 -e=di,ca" Queries discovery and cache events from node with '12345678' ID8. events Starts command in interactive mode.
Contains Visor command gc implementation.
Contains Visor command gc implementation.
+---------------------------------------------------------------------------+ | gc | Runs garbage collector on remote nodes. | | | If specific node is provided, garbage collector is run on that node. | | | Otherwise, it will be run on all nodes in topology. | +---------------------------------------------------------------------------+
gc
gc "{-id8=|-id=} {-c}" -id8=<node-id8>
ID8 of the node.
Note that either '-id8' or '-id' can be specified.
-id=<node-id>
ID of the node.
Note that either '-id8' or '-id' can be specified.
-c
Run DGC procedure on all caches.gc "-id8=12345678" Runs garbage collector on specified node. gc Runs garbage collector on all nodes in topology. gc "-id8=12345678 -c" Runs garbage collector and DGC procedure on all caches.
Contains Visor command kill implementation.
Contains Visor command kill implementation.
+--------------------------------+ | kill | Kills or restarts node. | +--------------------------------+
kill kill "-in|-ih" kill "{-r|-k} {-id8=|-id= }"
-in
Run command in interactive mode with ability to
choose a node to kill or restart.
Note that either '-in' or '-ih' can be specified.
This mode is used by default.
-ih
Run command in interactive mode with ability to
choose a host where to kill or restart nodes.
Note that either '-in' or '-ih' can be specified.
-r
Restart node mode.
Note that either '-r' or '-k' can be specified.
If no parameters provided - command starts in interactive mode.
-k
Kill (stop) node mode.
Note that either '-r' or '-k' can be specified.
If no parameters provided - command starts in interactive mode.
-id8=<node-id8>
ID8 of the node to kill or restart.
Note that either '-id8' or '-id' can be specified.
If no parameters provided - command starts in interactive mode.
-id=<node-id>
ID of the node to kill or restart.
Note that either '-id8' or '-id' can be specified.
If no parameters provided - command starts in interactive mode.kill
Starts command in interactive mode.
kill "-id8=12345678 -r"
Restart node with '12345678' ID8.
kill "-k"
Kill (stop) all nodes.
Contains Visor command license implementation.
Contains Visor command license implementation.
+---------------------------------------------------------------------------+ | license | Shows information about all licenses that are used on the grid. | | | Also can be used to update on of the licenses. | +---------------------------------------------------------------------------+
license
license "-f= -id=" -f=<path>
Path to new license XML file.
-id=<license-id>
ID of the license will be updated.license
Shows all licenses that are used on the grid.
license "-f=/path/to/new/license.xml -id=fbdea781-90e6-4d1b-b8b3-5b8c14aa2df7"
Copies new license file to all nodes that use license with provided ID.
Contains Visor command node implementation.
Contains Visor command node implementation.
+--------------------------------+ | node | Prints node statistics. | +--------------------------------+
node "{-id8=|-id=} {-a}"
node-id8=<node-id8>
ID8 of node. Either '-id8' or '-id' can be specified.
If neither specified - command starts in interactive mode.
-id=<node-id>
Full ID of node. Either '-id8' or '-id' can be specified.
If neither specified - command starts in interactive mode.
-a
Print extended information.
By default - only abbreviated statistics is printed.node
Starts command in interactive mode.
node "-id8=12345678"
Prints statistics for specified node.
node "-id8=12345678 -a"
Prints full statistics for specified node.
Visor 'ping' command implementation.
Visor 'ping' command implementation.
+--------------------+ | ping | Pings node. | +--------------------+
ping {"id81 id82 ... id8k"}id8k ID8 of the node to ping.
ping "12345678" Pings node with '12345678' ID8. ping Pings all nodes in the topology.
Contains Visor command start implementation.
Contains Visor command start implementation.
+-----------------------------------------------------+ | start | Starts one or more nodes on remote host(s). | | | Uses SSH protocol to execute commands. | +-----------------------------------------------------+
start "-f={-m= start "-h=} {-r}" {-p= } {-u= } {-pw= } {-k= } {-n= } {-g= } {-c= } {-s= } {-m= } {-r}"
-f=<path>
Path to INI file that contains topology specification.
-h=<hostname>
Hostname where to start nodes.
Can define several hosts if their IPs are sequential.
Example of range is 192.168.1.100~150,
which means all IPs from 192.168.1.100 to 192.168.1.150 inclusively.
-p=<num>
Port number (default is 22).
-u=<username>
Username (if not defined, current local username will be used).
-pw=<password>
Password (if not defined, private key file must be defined).
-k=<path>
Path to private key file. Define if key authentication is used.
-n=<num>
Expected number of nodes on the host.
If some nodes are started already, then only remaining nodes will be started.
If current count of nodes is equal to this number and '-r' flag is not set, then nothing will happen.
-g=<path>
Path to GridGain installation folder.
If not defined, GRIDGAIN_HOME environment variable must be set on remote hosts.
-c=<path>
Path to configuration file (relative to GridGain home).
If not provided, default GridGain configuration is used.
-s=<path>
Path to start script (relative to GridGain home).
Default is "bin/ggstart.sh" for Unix or
"bin\ggstart.bat" for Windows.
-m=<num>
Defines maximum number of nodes that can be started in parallel on one host.
This actually means number of parallel SSH connections to each SSH server.
Default is 5.
-r
Indicates that existing nodes on the host will be restarted.
By default, if flag is not present, existing nodes will be left as is.start "-h=10.1.1.10 -u=uname -pw=passwd -n=3" Starts three nodes with default configuration (password authentication). start "-h=192.168.1.100~104 -u=uname -k=/home/uname/.ssh/is_rsa -n=5" Starts 25 nodes on 5 hosts (5 nodes per host) with default configuration (key-based authentication). start "-f=start-nodes.ini -r" Starts topology defined in 'start-nodes.ini' file. Existing nodes are stopped.
Visor 'tasks' command implementation.
Visor 'tasks' command implementation.
+---------------------------------------------------------------------------------------+ | tasks | Prints statistics about tasks and executions. | | | | | | Note that this command depends on GridGain events. | | | | | | GridGain events can be individually enabled and disabled and disabled events | | | can affect the results produced by this command. Note also that configuration | | | of Event Storage SPI that is responsible for temporary storage of generated | | | events on each node can also affect the functionality of this command. | | | | | | By default - all events are enabled and GridGain stores last 10,000 local | | | events on each node. Both of these defaults can be changed in configuration. | +---------------------------------------------------------------------------------------+
tasks tasks "-l {-t=s|m|h|d} {-r}" tasks "-s={-t= tasks "-g {-t=s|m|h|d} {-r}" s|m|h|d} {-r}" tasks "-h {-t=s|m|h|d} {-r}" tasks "-n={-r}" tasks "-e="
-l
List all tasks and executions for a given time period.
Executions sorted chronologically (see '-r'), and tasks alphabetically
See '-t=<num>s|m|h|d' on how to specify the time limit.
Default time period is one hour, i.e '-t=1h'
This is a default mode when command is called without parameters.
-s=<substring>
List all tasks and executions for a given task name substring.
Executions sorted chronologically (see '-r'), and tasks alphabetically
See '-t=<num>s|m|h|d' on how to specify the time limit.
Default time period is one hour, i.e '-t=1h'
-g
List all tasks grouped by nodes for a given time period.
Tasks sorted alphabetically
See '-t=<num>s|m|h|d' on how to specify the time limit.
Default time period is one hour, i.e '-t=1h'
-h
List all tasks grouped by hosts for a given time period.
Tasks sorted alphabetically
See '-t=<num>s|m|h|d' on how to specify the time limit.
Default time period is one hour, i.e '-t=1h'
-t=<num>s|m|h|d
Defines time frame for '-l' parameter:
=<num>s Last <num> seconds.
=<num>m Last <num> minutes.
=<num>h Last <num> hours.
=<num>d Last <num> days.
-r
Reverse sorting of executions.
-n=<task-name>
Prints aggregated statistic for named task.
-e=<exec-id>
Prints aggregated statistic for given task execution.tasks "-l" Prints list of all tasks and executions for the last hour (default). tasks Prints list of all tasks and executions for the last hour (default). tasks "-l -t=5m" Prints list of tasks and executions that started during last 5 minutes. tasks "-s=Task" Prints list of all tasks and executions that have 'Task' in task name. tasks "-g" Prints list of tasks grouped by nodes. tasks "-g -t=5m" Prints list of tasks that started during last 5 minutes grouped by nodes. tasks "-h" Prints list of tasks grouped by hosts. tasks "-h -t=5m" Prints list of tasks that started during last 5 minutes grouped by hosts. tasks "-n=GridTask" Prints summary for task named 'GridTask'. tasks "-e=7D5CB773-225C-4165-8162-3BB67337894B" Traces task execution with ID '7D5CB773-225C-4165-8162-3BB67337894B'. tasks "-e=@e1" Traces task execution with ID taken from 'e1' memory variable.
Contains Visor command top implementation.
Contains Visor command top implementation.
+--------------------------------+ | top | Prints current topology. | +--------------------------------+
top "{-c1=e1 -c2=e2 ... -ck=ek} {-h= ... -h=} {-a}" -ck=ek<num>
This defines a mnemonic for node filter:
-cc Number of available CPUs on the node.
-cl Average CPU load (in %) on the node.
-aj Active jobs on the node.
-cj Cancelled jobs on the node.
-tc Thread count on the node.
-it Idle time on the node.
Note: <num> can have 's', 'm', or 'h' suffix indicating
seconds, minutes, and hours. By default (no suffix provided)
value is assumed to be in milliseconds.
-ut Up time on the node.
Note: <num> can have 's', 'm', or 'h' suffix indicating
seconds, minutes, and hours. By default (no suffix provided)
value is assumed to be in milliseconds.
-je Job execute time on the node.
-jw Job wait time on the node.
-wj Waiting jobs count on the node.
-rj Rejected jobs count on the node.
-hu Heap memory used (in MB) on the node.
-hm Heap memory maximum (in MB) on the node.
Comparison part of the mnemonic predicate:
=eq<num> Equal '=' to '' number.
=neq<num> Not equal '!=' to '' number.
=gt<num> Greater than '>' to '' number.
=gte<num> Greater than or equal '>=' to '' number.
=lt<num> Less than '<' to '' number.
=lte<num> Less than or equal '<=' to '' number.
-h=<host>
This defines a host to show nodes from.
Multiple hosts can be provided.
-a
This defines whether to show a separate table of nodes
with detail per-node information.top "-cc=eq2" Prints topology for all nodes with two CPUs. top "-cc=eq2 -a" Prints full information for all nodes with two CPUs. top "-h=10.34.2.122 -h=10.65.3.11" Prints topology for provided hosts. top Prints full topology.
Contains Visor command vvm implementation.
Contains Visor command vvm implementation.
+-----------------------+ | vvm | Opens VisualVM. | +-----------------------+
vvm "{-home=dir} {-id8=} {-id=}" -home=dir
VisualVM home directory.
If not specified, PATH and JAVA_HOME will be searched
-id8=<node-id8>
ID8 of node.
Either '-id8' or '-id' can be specified.
-id=<node-id>
Full ID of node.
Either '-id8' or '-id' can be specified.vvm "-id8=12345678" Opens VisualVM connected to JVM for node with '12345678' ID8. vvm "-id=5B923966-85ED-4C90-A14C-96068470E94D" Opens VisualVM connected to JVM for node with given full node ID. vvm "-home=C:\VisualVM -id8=12345678" Opens VisualVM installed in 'C:\VisualVM' directory for specified node. vvm Opens VisualVM connected to all nodes.
Overview
Provides
ASCII-based table with minimal styling support.