input
stringlengths
10
828
output
stringlengths
5
107
public void run log info starting data node in data start data xceive server data xceive server start while should run try offer service catch exception ex log error exception string utils stringify exception ex if should run try thread sleep 5000 catch interrupted exception ie wait for data xceive server to terminate try this data xceive server join catch interrupted exception ie log info finishing data node in data
no matter what kind of exception we get keep retrying to offer service
public boolean rename utf8 src utf8 dst throws ioexception check open return namenode rename src to string dst to string
make a direct connection to namenode and manipulate structures
public boolean delete utf8 src throws ioexception check open return namenode delete src to string
make a direct connection to namenode and manipulate structures
public file get fs image name throws ioexception return namesystem get fs image name
returns the name of the fs image file
public file get fs edit name throws ioexception return namesystem get fs edit name
returns the name of the edits file
public void localize configuration job conf conf throws ioexception super localize configuration conf conf set num map tasks num maps
localize the given job conf to be specific for this task
private void purge task task in progress tip throws ioexception if tip null log info about to purge task tip get task get task id remove the task from running jobs removing the job if it s the last task remove task from job tip get task get job id tip tip job has finished
remove the tip and update all relevant state
private datanode info best node datanode info nodes tree set dead nodes throws ioexception if nodes null for int i 0 i nodes length i if dead nodes contains nodes i return nodes i throw new ioexception no live nodes contain current block
pick the best node from which to stream the data
private void add to invalidates block b datanode info n collection block invalidate set recent invalidate sets get n get storage id if invalidate set null invalidate set new array list block recent invalidate sets put n get storage id invalidate set invalidate set add b
adds block to list of blocks which will be invalidated on
private void remove marked tasks string task tracker purge all the marked tasks which were running at task tracker tree set string marked task set tree set string tracker to marked tasks map get task tracker if marked task set null for string taskid marked task set remove task entry taskid log info removed completed task taskid from task tracker clear tracker to marked tasks map remove task tracker
remove all marked tasks running on a given
public void update metrics counters counters get counters for string group name counters get group names counters group group counters get group group name job metrics set tag group group get display name for string counter group get counter names long value group get counter counter job metrics set tag counter group get display name counter job metrics set metric value float value job metrics update
called periodically by job tracker metrics to update the metrics for
private void kill overflowing tasks throws ioexception if enough free space min space kill accept new tasks false we give up do not accept new tasks until all the ones running have finished and they re all cleared up synchronized this task in progress kill me find task to kill if kill me null string msg tasktracker running out of space killing task log info kill me get task get task id msg kill me report diagnostic info msg purge task kill me
check if were dangerously low on disk space
private boolean enough free space long min space throws ioexception if min space 0 return true string local dirs f conf get local dirs for int i 0 i local dirs length i df df null if local dirs df contains key local dirs i df df local dirs df get local dirs i else df new df new file local dirs i f conf local dirs df put local dirs i df if df get available min space return false return true
check if all of the local directories have enough
public void run while true try map output location loc null long size 1 synchronized scheduled copies while scheduled copies is empty scheduled copies wait loc map output location scheduled copies remove 0 try start loc size copy output loc catch ioexception e log warn reduce task get task id copy failed loc get map task id from loc get host log warn string utils stringify exception e finally finish size catch interrupted exception e return all done catch throwable th log error map output copy failure string utils stringify exception th
loop forever and fetch map outputs as they become available
public void set map output compression type sequence file compression type style set map output compression type style to string
set the compression type for the map outputs
public void set map output compressor class class codec class set compress map output true set class mapred output compression codec codec class compression codec class
set the given class as the compression codec for the map outputs
public int get max tasks return max tasks
the maximum capacity for running tasks in the cluster
public class get map output compressor class class default value string name get mapred output compression codec if name null return default value else try return get class by name name catch class not found exception e throw new illegal argument exception compression codec name was not found e
get the codec for compressing the map outputs
public vector report tasks in progress boolean should be map boolean should be complete vector results new vector task in progress tips null if should be map tips maps else tips reduces for int i 0 i tips length i if tips i is complete should be complete results add tips i return results
return a treeset of completed task in progress objects
public class get map output key class class retv get class mapred mapoutput key class null writable comparable class if retv null retv get output key class return retv
get the key class for the map output data
public void set map output key class class the class set class mapred mapoutput key class the class writable comparable class
set the key class for the map output data
public counters get counters return counters sum get map counters get reduce counters
returns the total job counters by adding together the map and the
public class get map output value class class retv get class mapred mapoutput value class null writable class if retv null retv get output value class return retv
get the value class for the map output data
private counters sum task counters task in progress tips counters counters new counters for task in progress tip tips counters incr all counters tip get counters return counters
returns a counters instance representing the sum of all the counters in
public boolean expired hard limit if now last update lease hardlimit period return true return false
returns true if the hard limit timer has expired
public void set map output value class class the class set class mapred mapoutput value class the class writable class
set the value class for the map output data
public void close throws ioexception get task get progress set status closed this map output file remove all get task get task id
delete all of the temporary map output files
public boolean expired soft limit if now last update lease softlimit period return true return false
returns true if the soft limit timer has expired
public counters get counters throws ioexception return job submit client get job counters get job id
returns the counters for this job
public string get display name return localize counter group name group name
returns localized name of the group
public string get display name string counter return localize counter name counter
returns localized name of the specified counter
public long get counter string counter long result group counters get counter return result null 0 l result
returns the value of the specified counter or 0 if the counter does
public int size return group counters size
returns the number of counters in this group
public datanode info get current datanode return dfsinput stream in stream get current datanode
returns the datanode from which the stream is currently reading
public void log log log log info counters size collection string group names get group names for string group name group names group group get group group name log info group get display name for string counter name group get counter names log info group get display name counter name group get counter counter name
logs the current counter values
public block get current block return dfsinput stream in stream get current block
returns the block containing the target position
public task in progress get task in progress string tipid for int i 0 i maps length i if tipid equals maps i get tipid return maps i for int i 0 i reduces length i if tipid equals reduces i get tipid return reduces i return null
return the task in progress that matches the tipid
public boolean get speculative execution return get boolean mapred speculative execution true
should speculative execution be used for this job
public void remove all string task id throws ioexception conf delete local files task id
removes all of the files related to a task
public void cleanup storage throws ioexception conf delete local files
removes all contents of temporary storage
public void set speculative execution boolean new val set boolean mapred speculative execution new val
turn on or off speculative execution for this job
public boolean prepare throws ioexception if super prepare return false this map output file remove all get task get task id return true
delete any temporary files from previous failed attempts
public void close throws ioexception log info get task done removing files this map output file remove all get task get task id
delete all of the temporary map output files
public void set run state int state this run state state
change the current run state of the job
public long get total log size throws ioexception if initialized init return log file size
return the total logical log size written by the task including
public raw key value iterator sort merge sort m new merge sort this int count super count if count 0 return null int pointers int super pointers int pointers copy new int count system arraycopy pointers 0 pointers copy 0 count m merge sort pointers pointers copy 0 count return new mrsort result iterator super key val buffer pointers copy super start offsets super key lengths super value lengths
the sort method derived from basic type sorter base and overridden here
public int compare int writable i int writable j return super compare i get j get
the implementation of the compare method from comparator
public string get job id return jobid
get the job id
public url get url try return new url url to string catch ioexception ie return null
get the link to the web ui for details of the job
private string make unique string string unique base string buffer result new string buffer result append unique base if is map task result append m else result append r result append id format format partition return result to string
make a unique name for this tip
public void set mapred job id string mapred job id this job id mapred job id
set the mapred id for this job
public string get tipid return this id
return an id for this task not its component taskid threads
public boolean is map task return split null
whether this is a map task
public boolean is complete return completes 0
is this tip complete
public boolean is complete string taskid task status status task status task statuses get taskid if status null return false return completes 0 status get run state task status state succeeded
is the given taskid in this tip complete
private void check running state running job running null try running jc get job this mapred job id if running is complete if running is successful this state job success else this state job failed this message job failed try running kill job catch ioexception e1 try this jc close catch ioexception e2 catch ioexception ioe this state job failed this message string utils stringify exception ioe try running kill job catch ioexception e1 try this jc close catch ioexception e1
check the state of this running job
public void submit try if the job conf get boolean create empty dir if nonexist false file system fs file system get the job conf path input paths the job conf get input paths for int i 0 i input paths length i if fs exists input paths i try fs mkdirs input paths i catch ioexception e running job running jc submit job the job conf this mapred job id running get job id this state job running catch ioexception ioe this state job failed this message string utils stringify exception ioe
submit this job to mapred
public void commit throws ioexception for path f path final name to file info key set commit f path false safe to clear map now final name to file info clear
commits files to their final locations as passed in create methods
protected void set phase task status phase p this phase p
set current phase of the task
public void abort path p throws ioexception abort p true
aborts a single file
public void run try map pair to output multithreaded map runner this mapper map key value output reporter catch ioexception ex if there is an ioexception during the call it is set in an instance variable of the multithreaded map runner from where it will be rethrown synchronized multithreaded map runner this if multithreaded map runner this io exception null multithreaded map runner this io exception ex
executes a mapper
public void localize configuration job conf conf throws ioexception conf set mapred tip id tip id conf set mapred task id task id conf set boolean mapred task is map is map task conf set int mapred task partition partition conf set mapred job id job id
localize the given job conf to be specific for this task
public void abort throws ioexception for path f path final name to file info key set abort f path false safe to clean now final name to file info clear
aborts the file creation all uncommitted files created by this phased file system
public boolean should close for closed job string taskid if the thing has never been closed and it belongs to this tip and this tip is somehow finished then true task status ts task status task statuses get taskid if ts null tasks reported closed contains taskid job get status get run state job status running tasks reported closed add taskid return true else if is map task is complete tasks reported closed contains taskid tasks reported closed add taskid return true else return false
returns whether a component task thread should be
private task in progress get tip string jobid string tipid job in progress job job in progress jobs get jobid return job null null task in progress job get task in progress tipid
returns specified task in progress or null
public void completed string taskid record that this taskid is complete completed task taskid now that the tip is complete the other speculative subtasks will be closed when the owning tasktracker reports in and calls should close on this object this completes recompute progress
indicate that one of the taskids in this task in progress
public task status get task status string taskid return task statuses get taskid
get the status of the specified task
private void internal release create utf8 src utf8 holder file under construction v pending creates remove src if v null name node state change log debug dir name system internal release create src is removed from pending creates for holder failure for iterator block it2 v get blocks iterator it2 has next block b it2 next pending create blocks remove b else name node state change log warn dir name system internal release create attempt to release a create lock on src to string that was not in peding creates
release a pending file creation lock
public boolean has failed on machine string tracker return machines where failed contains tracker
has this task already failed on this machine
public boolean has run on machine string tracker return this active tasks values contains tracker has failed on machine tracker
was this task ever scheduled to run on this machine
public void renew lease utf8 holder throws ioexception synchronized leases if is in safe mode throw new safe mode exception cannot renew lease for holder safe mode lease lease leases get holder if lease null sorted leases remove lease lease renew sorted leases add lease
renew the lease s held by the given client
public string get registration id return ns integer to string dir namespace id
get registration id for datanodes based on the namespace id
private void proccess over replicated block block block short replication collection datanode descriptor containing nodes blocks map get block if containing nodes null return collection datanode descriptor non excess new array list datanode descriptor for iterator datanode descriptor it containing nodes iterator it has next datanode descriptor cur it next collection block excess blocks excess replicate map get cur get storage id if excess blocks null excess blocks contains block if cur is decommission in progress cur is decommissioned non excess add cur choose excess replicates non excess block replication
find how many of the containing nodes are extra if any
public long total capacity synchronized heartbeats return total capacity
total raw bytes
public long total remaining synchronized heartbeats return total remaining
total non used raw bytes
public int total load synchronized heartbeats return total load
total number of connections
private boolean check decommission state internal datanode descriptor node check to see if there are all blocks in this decommisioned node has reached their target replication factor if node is decommission in progress if is replication in progress node node set decommissioned log info decommission complete for node node name if node is decommissioned return true return false
change if appropriate the admin state of a datanode to
public void cleanup throws ioexception close if preserve input fs delete segment path name
the default cleanup
public datanode descriptor get datanode by name string name throws ioexception for iterator datanode descriptor it datanode map values iterator it has next datanode descriptor node it next if node get name equals name return node return null
find data node by its name
private datanode info get datanode by index int index int i 0 for iterator datanode descriptor it datanode map values iterator it has next datanode info node it next if i index return node i return null
stop at and return the datanode at index used for content browsing
public boolean is neighbor dt triangle triangle int count 0 for pnt vertex this if triangle contains vertex count return count 1
true iff triangles are neighbors
public triangulation dt triangle triangle tri graph new graph dt triangle tri graph add triangle most recent triangle
all sites must fall within the initial triangle
public dt triangle neighbor opposite pnt site dt triangle triangle if triangle contains site throw new illegal argument exception bad vertex not in triangle for dt triangle neighbor tri graph neighbors triangle if neighbor contains site return neighbor return null
report neighbor opposite the given vertex of triangle
public void delaunay place pnt site uses straightforward scheme rather than best asymptotic time locate containing triangle dt triangle triangle locate site give up if no containing triangle or if site is already in dt if triangle null throw new illegal argument exception no containing triangle if triangle contains site return determine the cavity and update the triangulation set dt triangle cavity get cavity site triangle most recent update site cavity
place a new site into the dt
private int find max2 d int list int size1 int val list 0 0 for int i 0 i size1 i for int j 0 j size1 j if list i j val val list i j return val
maximum of an 2 d array
public void evaluate cds for int i 0 i nodes array size i nodes array get i set dominator has two unconnected neighbors nodes array get i if is all unmarked as dominator nodes array get nodes array size 1 set dominator true else prune rule1 prune rule2 init dominator neighborhood attach dominees to closest dominators
evaluates minimum connected dominating set of the run
private boolean is all unmarked as dominator for int i 0 i nodes array size i if nodes array get i is dominator return false return true
the algorithm we used to calculate cds is a distributed approach
private void init dominator neighborhood array list gateway dominator list new array list gateway gateway tmp for int k 0 k nodes array size k if nodes array get k is dominator tmp nodes array get k dominator list add tmp for int i 0 i dominator list size i for int j 0 j dominator list size j if i j gateway ptr1 ptr1 dominator list get i gateway ptr2 ptr2 dominator list get j if is neighbor ptr1 ptr2 dominator list get i get neighbor dominator list add dominator list get j
connects neighbor dominators
public boolean add message message m return add message m false
if the inbox contains the message m ignore it
public message get next unread message if ptr message messages size message message messages get ptr message message mark as read ptr message return message else return null
retrieves next unread message
public void empty messages clear
clears the message box
private int find numberof cut vertices int result 0 for int i 0 i relay node array size i gateway gateway relay node array get i if is cut vertex gateway result return result
only looks at the relay network
private void set edge pruned hash map edge integer pruned edge edge for edge edge1 pruned key set if edge u edge1 u edge v edge1 v edge u edge1 v edge v edge1 u pruned put edge1 1
old version but still has some usages
public void add url filter iurl filter filter f includes filters add filter
adds a new filter to this composite
public void remove url filter iurl filter filter f includes filters remove filter
removes the given filter from this composite
protected void copy children file dir throws ioexception file array dir list files if array null iwrf resource provider provider f resource get provider page link path f resource get path string type f resource get type for file child array page link child path path append path child get name iwrf resource child resource provider get resource type child path file adapter child adapter child resource get adapter file adapter class child adapter copy from child true
recursively copies all children of the specified directory in
private void set metadata file file throws ioexception iproperty adapter properties f resource get adapter iproperty adapter class long date file last modified string formatted date date util format date date properties set property last modified formatted date
copies the metadata about the specified file in the properties of this
public void write xhtml output stream out throws ioexception final output stream writer writer new output stream writer out utf 8 write xhtml1 writer
writes out the xhtml representation of the underlying resource in the
protected string filter uri site url uri return uri to string
this method filters the representation of the given url and returns it
private void normalize attributes named node map attributes if attributes null return for int i attributes get length 1 i 0 i attr attr attr attributes item i string name attr get name name name to lower case if style equals name name starts with on attributes remove named item name
normalizes attributes from the given map