org.apache.hadoop.hbase.master

Class CachedClusterId

java.lang.Object

org.apache.hadoop.hbase.master.CachedClusterId

@InterfaceAudience.Private
public class CachedClusterId
extends Object

Caches the cluster ID of the cluster. For standby masters, this is used to serve the client RPCs that fetch the cluster ID. ClusterID is only created by an active master if one does not already exist. Standby masters just read the information from the file system. This class is thread-safe. TODO: Make it a singleton without affecting concurrent junit tests.

Field Summary

Fields

Modifier and Type	Field and Description
private AtomicInteger	cacheMisses
private ClusterId	clusterId
(package private) AtomicBoolean	fetchInProgress
private org.apache.hadoop.fs.FileSystem	fs
private AtomicBoolean	isClusterIdSet
static org.slf4j.Logger	LOG
private static int	MAX_FETCH_TIMEOUT_MS
private org.apache.hadoop.fs.Path	rootDir
private Server	server

Constructor Summary

Constructors

Constructor and Description
CachedClusterId(Server server, org.apache.hadoop.conf.Configuration conf)

Method Summary

Modifier and Type	Method and Description
private boolean	attemptFetch() Attempts to fetch the cluster ID from the file system.
int	getCacheStats()
private String	getClusterId() Returns a cached copy of the cluster ID.
String	getFromCacheOrFetch() Fetches the ClusterId from FS if it is not cached locally.
private void	setClusterId(ClusterId id) Succeeds only once, when setting to a non-null value.
private void	waitForFetchToFinish()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

LOG

public static final org.slf4j.Logger LOG

MAX_FETCH_TIMEOUT_MS

private static final int MAX_FETCH_TIMEOUT_MS

See Also:

Constant Field Values

rootDir

private final org.apache.hadoop.fs.Path rootDir

fs

private final org.apache.hadoop.fs.FileSystem fs

fetchInProgress

AtomicBoolean fetchInProgress

isClusterIdSet

private AtomicBoolean isClusterIdSet

clusterId

private ClusterId clusterId

server

private final Server server

cacheMisses

private AtomicInteger cacheMisses

Constructor Detail

CachedClusterId

public CachedClusterId(Server server,
                       org.apache.hadoop.conf.Configuration conf)
                throws IOException

Throws:

IOException

Method Detail

setClusterId

private void setClusterId(ClusterId id)

Succeeds only once, when setting to a non-null value. Overwrites are not allowed.

getClusterId

private String getClusterId()

Returns a cached copy of the cluster ID. null if the cache is not populated.

attemptFetch

private boolean attemptFetch()

Attempts to fetch the cluster ID from the file system. If no attempt is already in progress, synchronously fetches the cluster ID and sets it. If an attempt is already in progress, returns right away and the caller is expected to wait for the fetch to finish.

Returns:

true if the attempt is done, false if another thread is already fetching it.

waitForFetchToFinish

private void waitForFetchToFinish()
                           throws InterruptedException

Throws:

InterruptedException

getFromCacheOrFetch

public String getFromCacheOrFetch()

Fetches the ClusterId from FS if it is not cached locally. Atomically updates the cached copy and is thread-safe. Optimized to do a single fetch when there are multiple threads are trying get from a clean cache.

Returns:

ClusterId by reading from FileSystem or null in any error case or cluster ID does not exist on the file system or if the server initiated a tear down.

getCacheStats

public int getCacheStats()