http://qizmt.googlecode.com/svn/wiki/images/Qizmt_logo_small.png' alt='Qizmt logo (small)' />
Back to Qizmt Tutorial or Wiki Main
Qizmt Command Line
Usage Descriptionretrylogmd5 <logfile>
writes all output to the logfile
edit <jobs.xml>
edit the specified jobs source code XML file
exec ["</xpath>=<value>"] <jobs.xml>
run the jobs source code XML file
addmachine <host>
add a machine to the cluster
removemachine <host>
remove a machine from the cluster
ps
distributed processes, schedule and queue info
who
show who is logged on
history [<count>]
show command history
killall
kill all jobs, clean any orphaned intermediate data
gen <output-dfsfile>
<outputsize>
[type=<bin|ascii|word>]
[row=<size>]
[writers=<count>]
[rand=custom]
generate random binary, ascii or word data
combine <inputfiles...> [+ <outputfile>]
combine files into one
format machines=localhost
format machines=localhost
info [<dfspath>[:<host>]]
[-s short host name]
[-mt multi-threaded]
information for DFS or a DFS file
head <dfspath>[:<host>:<part>] [<count>]
show first few lines of file
put <netpath> [<dfspath>[@<recordlen>]]
put a file into DFS
fput files|dirs=<item[,item,item]>|@<filepath to list>
[pattern=<pattern>]
[mode=continuous]
[dfsfilename=<targetfilename>]
put files into DFS
putbinary <wildcard> <dfspath>
put binary into DFS
get [parts=<first>[-[<last>]]] <dfspath> <netpath>
get a file from DFS
fget <dfspath>
<targetFolder>[ <targetFolder> <targetFolder>] |
@<filepath to target folder list> [-gz] [-md5]
get a file from DFS
getbinary <dfspath> <netpath>
get binary from DFS
del <dfspath|wildcard>
delete a DFS file using multiple threads
rename <dfspath-old> <dfspath-new>
rename a DFS file
getjobs <netpath.dj>
archive all jobs
putjobs <netpath.dj>
import jobs archive
ls
DFS file listing
countparts <dfspath>
get parts count of a file
invalidate <cacheName> <fileNodeName>
invalidate a file node of the cache
health [-a check DFS health]
[-v verify driver]
[-mt multi-threaded]
Show the health of the machines in the cluster
examples
generate example jobs source code
importdir <netpath>
import jobs into DFS
listinstalldir
List all installed directories
harddrivespeedtest [<filesize>]
Test write/read hard drive speed
networkspeedtest [<filesize>]
Test upload/download network speed test
exechistory
List the most recent executed commands
cputemp
List cpu temperature
ghost
List ghost data files
perfmon <network|cputime|diskio|availablememory>
[a=<Number of readings to get. Return average.>]
[t=<Number of threads>]
[s=<Milliseconds of sleep to take between readings>]
get Perfmon counter readings
perfmon generic
o=<Object/category name>
c=<Counter name>
i=<Instance Name>
[f Display readings in friendly byte size units]
[a=<Number of readings to get. Return average.>]
[t=<Number of threads>]
[s=<Milliseconds of sleep to take between readings>]
specify a Perfmon counter to read
packetsniff [t=<Number of threads>]
[s=<Milliseconds to sniff>]
[v verbose]
[a include non-cluster machines]
Sniff packets
md5 <dfsfile>
compute MD5 of DFS data file
checksum <dfsfile>
compute sum of DFS data file
sorted <dfsfile>
check if a DFS data file has sorted lines
nearprime <positiveNum>
find the nearest prime number
genhostnames <pattern> <startNum>
<endNum> [<delimiter>]
generate host names
viewlogs [machine=<machineName>]
[count=<number of entries to return>]
view log entries
clearlogs
clear logs from all machines in the cluster
maxuserlogsview
view maxUserLogs configuration
maxuserlogsupdate <integer>
update maxUserLogs configuration
maxdglobalsview
view maxDGlobals configuration
maxdglobalsupdate <integer>
update maxDGlobals configuration
recordsize <user-size>
returns bytes of user-friendly size
swap <file1> <file2>
file names to swap
regressiontest basic
basic regression test Qizmt
kill <JobID>
kill the specified Job Identifier
enqueue command=<value>
[ExecTimeout=<secs> Maximum seconds Qizmt exec can run
OnTimeout=<tcmd>] Run on timeout; e.g. Qizmt kill #JID#
Adds a command to the end of the queue
queuekill <QueueID>
Removes the specified Queue Identifier
clearqueue
Removes all entries from the queue
schedule command=<value>
start=<now|<datetime>>
[frequency=<seconds>]
[texceptions=<<datetime>[-<datetime>]>[,...]] ranges when not to run
[wexceptions=<weekday>[,...]] whole weekdays not to run
[wtexceptions=<wdtime>[,...]] time on day-of-week not to run
adds a command entry to the scheduler
(datetime format is [M[/D[/Y]].][h:m[:s][AM|PM]]
)
(wdtime format is <weekday>@<h:m[:s][AM|PM]>[-<h:m[:s][AM|PM]>]
)
pauseschedule <ScheduleID>
Pauses the specified Schedule Identifier
unpauseschedule <ScheduleID>
Un-pauses the specified Schedule Identifier
unschedule <ScheduleID>
Removes the specified Schedule Identifier
clearschedule
Removes all entries from the scheduler
shuffle <source> <target>
Shuffle underlying parts of a rectangular binary file, maintaining chunk order
spread <dfsfile> <out-dfsfile>
Spread a DFS file across the cluster