STABLE

java.lang.Object
- com.fr.third.jgroups.stack.Protocol
- - com.fr.third.jgroups.protocols.pbcast.STABLE

```
public class STABLE
extends Protocol
```
Computes the broadcast messages that are stable; i.e., have been delivered by all members. Sends STABLE events down the stack when this is the case. This allows NAKACK{2,3} to garbage collect messages that have been seen by all members.
Works as follows: periodically (desired_avg_gossip) or when having received a number of bytes (max_bytes), every member sends its digest (highest seqno delivered, received) to the cluster (send_stable_msgs_to_coord_only=false) or the current coordinator (send_stable_msgs_to_coord_only=true).
The recipient updates a stability vector, which maintains the highest seqno delivered/receive for each member and initially contains no data, when such a message is received.
When messages from all members have been received, a stability message is mcast, which causes all members to send a STABLE event down the stack (triggering garbage collection in the NAKACK{2,3} layer).
When send_stable_msgs_to_coord_only is true, far fewer messages are exchanged, as members don't multicast STABLE messages, but instead send them only to the coordinator.

作者:

Bela Ban

嵌套类概要

嵌套类
限定符和类型	类和说明
`protected class`	`STABLE.ResumeTask`
`protected class`	`STABLE.StabilitySendTask` Multicasts a STABILITY message
`static class`	`STABLE.StableHeader`
`protected class`	`STABLE.StableTask` Mcast periodic STABLE message.

字段概要

字段
限定符和类型	字段和说明
`protected double`	`cap` 已过时。
`protected Address`	`coordinator`
`protected long`	`desired_avg_gossip` Sends a STABLE gossip every 20 seconds on average. 0 disables gossiping of STABLE messages
`protected MutableDigest`	`digest`
`protected boolean`	`initialized`
`protected Address`	`local_addr`
`protected java.util.concurrent.locks.Lock`	`lock`
`protected long`	`max_bytes` Total amount of bytes from incoming messages (default = 0 = disabled).
`protected static long`	`MAX_SUSPEND_TIME`
`protected long`	`num_bytes_received` The total number of bytes received from unicast and multicast messages
`protected int`	`num_stability_msgs_received`
`protected int`	`num_stability_msgs_sent`
`protected int`	`num_stable_msgs_received`
`protected int`	`num_stable_msgs_sent`
`protected java.util.concurrent.locks.Lock`	`received`
`protected java.util.concurrent.Future<?>`	`resume_task_future`
`protected java.lang.Object`	`resume_task_mutex`
`protected boolean`	`send_stable_msgs_to_coord_only`
`protected long`	`stability_delay` delay before we send STABILITY msg (give others a change to send first).
`protected java.util.concurrent.locks.Lock`	`stability_lock`
`protected java.util.concurrent.Future<?>`	`stability_task_future`
`protected java.util.concurrent.Future<?>`	`stable_task_future`
`protected java.util.concurrent.locks.Lock`	`stable_task_lock`
`protected boolean`	`suspended` When true, don't take part in garbage collection: neither send STABLE messages nor handle STABILITY messages
`protected TimeScheduler`	`timer`
`protected View`	`view`
`protected FixedSizeBitSet`	`votes` Keeps track of who we already heard from (STABLE_GOSSIP msgs).

从类继承的字段 com.fr.third.jgroups.stack.Protocol
down_prot, ergonomics, id, log, name, stack, stats, up_prot

构造器概要

构造器
构造器和说明

STABLE()

构造器
构造器和说明
`STABLE()`

方法概要

方法
限定符和类型	方法和说明
`protected boolean`	`addVote(int rank)` Adds mbr to votes and returns true if we have all the votes, otherwise false.
`protected static boolean`	`allVotesReceived(FixedSizeBitSet votes)` Votes is already locked and guaranteed to be non-null
`java.lang.Object`	`down(Event evt)` An event is to be sent down the stack.
`void`	`gc()`
`long`	`getBytes()`
`long`	`getDesiredAverageGossip()`
`protected Digest`	`getDigest()`
`long`	`getMaxBytes()`
`protected static int`	`getRank(Address member, View v)`
`int`	`getStabilityReceived()`
`int`	`getStabilitySent()`
`int`	`getStableReceived()`
`int`	`getStableSent()`
`boolean`	`getStableTaskRunning()`
`protected void`	`handleRegularMessage(Message msg)`
`protected void`	`handleStabilityMessage(Digest stable_digest, Address sender, ViewId view_id)`
`protected void`	`handleStableMessage(Digest d, Address sender, ViewId view_id)` Digest d contains (a) the highest seqnos deliverable for each sender and (b) the highest seqnos seen for each member.
`protected void`	`handleUpEvent(STABLE.StableHeader hdr, Address sender, Digest digest)`
`protected void`	`handleViewChange(View v)`
`void`	`init()` Called after instance has been created (null constructor) and before protocol is started.
`static Buffer`	`marshal(Digest digest)`
`protected java.lang.String`	`printDigest(Digest digest)`
`protected Digest`	`readDigest(byte[] buffer, int offset, int length)`
`java.util.List<java.lang.Integer>`	`requiredDownServices()` List of events that are required to be answered by some layer below
`protected void`	`resetDigest()`
`protected void`	`resetNumBytes()`
`void`	`resetStats()`
`protected void`	`resume()`
`protected void`	`sendStabilityMessage(Digest tmp, ViewId view_id)` Schedules a stability message to be mcast after a random number of milliseconds (range 1-5 secs).
`protected void`	`sendStableMessage(boolean send_in_background)` Broadcasts a STABLE message of the current digest to all members (or the coordinator only).
`void`	`setDesiredAverageGossip(long gossip_interval)`
`void`	`setMaxBytes(long max_bytes)`
`void`	`start()` This method is called on a `Channel.connect(String)`.
`protected void`	`startResumeTask(long max_suspend_time)`
`protected void`	`startStabilityTask(Digest d, ViewId view_id, long delay)`
`protected void`	`startStableTask()`
`void`	`stop()` This method is called on a `Channel.disconnect()`.
`protected void`	`stopResumeTask()`
`protected void`	`stopStabilityTask()`
`protected void`	`stopStableTask()`
`protected void`	`suspend(long timeout)`
`java.lang.Object`	`up(Event evt)` An event was received from the layer below.
`void`	`up(MessageBatch batch)` Sends up a multiple messages in a `MessageBatch`.
`protected void`	`updateLocalDigest(Digest d, Address sender)` Update my own digest from a digest received by somebody else.

从类继承的方法 com.fr.third.jgroups.stack.Protocol
accept, destroy, dumpStats, enableStats, getConfigurableObjects, getDownProtocol, getDownServices, getId, getIdsAbove, getLevel, getName, getProtocolStack, getSocketFactory, getThreadFactory, getTransport, getUpProtocol, getUpServices, getValue, isErgonomics, level, parse, printStats, providedDownServices, providedUpServices, requiredUpServices, resetStatistics, setDownProtocol, setErgonomics, setId, setLevel, setProtocolStack, setSocketFactory, setUpProtocol, setValue, setValues, statsEnabled

从类继承的方法 java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- 字段详细资料
  - MAX_SUSPEND_TIME
```
protected static final long MAX_SUSPEND_TIME
```
    另请参阅:
    常量字段值
  - desired_avg_gossip
```
protected long desired_avg_gossip
```
    Sends a STABLE gossip every 20 seconds on average. 0 disables gossiping of STABLE messages
  - stability_delay
```
protected long stability_delay
```
    delay before we send STABILITY msg (give others a change to send first). This should be set to a very small number (> 0 !) if max_bytes is used
  - max_bytes
```
protected long max_bytes
```
    Total amount of bytes from incoming messages (default = 0 = disabled). When exceeded, a STABLE message will be broadcast and num_bytes_received reset to 0 . If this is > 0, then ideally stability_delay should be set to a low number as well
  - cap
```
@Deprecated
protected double cap
```
    已过时。
  - send_stable_msgs_to_coord_only
```
protected boolean send_stable_msgs_to_coord_only
```
  - num_stable_msgs_sent
```
protected int num_stable_msgs_sent
```
  - num_stable_msgs_received
```
protected int num_stable_msgs_received
```
  - num_stability_msgs_sent
```
protected int num_stability_msgs_sent
```
  - num_stability_msgs_received
```
protected int num_stability_msgs_received
```
  - local_addr
```
protected Address local_addr
```
  - view
```
protected volatile View view
```
  - digest
```
protected volatile MutableDigest digest
```
  - votes
```
protected FixedSizeBitSet votes
```
    Keeps track of who we already heard from (STABLE_GOSSIP msgs). This is all 0's, and we set the sender when a STABLE message is received. When the bitset is all 1's (responses from all members), we send a STABILITY message
  - lock
```
protected final java.util.concurrent.locks.Lock lock
```
  - stability_task_future
```
protected java.util.concurrent.Future<?> stability_task_future
```
  - stability_lock
```
protected final java.util.concurrent.locks.Lock stability_lock
```
  - stable_task_future
```
protected java.util.concurrent.Future<?> stable_task_future
```
  - stable_task_lock
```
protected final java.util.concurrent.locks.Lock stable_task_lock
```
  - timer
```
protected TimeScheduler timer
```
  - num_bytes_received
```
protected long num_bytes_received
```
    The total number of bytes received from unicast and multicast messages
  - received
```
protected final java.util.concurrent.locks.Lock received
```
  - suspended
```
protected volatile boolean suspended
```
    When true, don't take part in garbage collection: neither send STABLE messages nor handle STABILITY messages
  - initialized
```
protected boolean initialized
```
  - resume_task_future
```
protected java.util.concurrent.Future<?> resume_task_future
```
  - resume_task_mutex
```
protected final java.lang.Object resume_task_mutex
```
  - coordinator
```
protected volatile Address coordinator
```
- 构造器详细资料
  - STABLE
```
public STABLE()
```
- 方法详细资料
  - getDesiredAverageGossip
```
public long getDesiredAverageGossip()
```
  - setDesiredAverageGossip
```
public void setDesiredAverageGossip(long gossip_interval)
```
  - getMaxBytes
```
public long getMaxBytes()
```
  - setMaxBytes
```
public void setMaxBytes(long max_bytes)
```
  - getBytes
```
public long getBytes()
```
  - getStableSent
```
public int getStableSent()
```
  - getStableReceived
```
public int getStableReceived()
```
  - getStabilitySent
```
public int getStabilitySent()
```
  - getStabilityReceived
```
public int getStabilityReceived()
```
  - getStableTaskRunning
```
public boolean getStableTaskRunning()
```
  - resetStats
```
public void resetStats()
```
    覆盖:
    
    resetStats 在类中 Protocol
  - requiredDownServices
```
public java.util.List<java.lang.Integer> requiredDownServices()
```
    从类复制的说明: Protocol
    
    List of events that are required to be answered by some layer below
    
    覆盖:
    
    requiredDownServices 在类中 Protocol
  - suspend
```
protected void suspend(long timeout)
```
  - resume
```
protected void resume()
```
  - init
```
public void init()
          throws java.lang.Exception
```
    从类复制的说明: Protocol
    
    Called after instance has been created (null constructor) and before protocol is started. Properties are already set. Other protocols are not yet connected and events cannot yet be sent.
    
    覆盖:
    
    init 在类中 Protocol
    
    抛出:
    
    java.lang.Exception - Thrown if protocol cannot be initialized successfully. This will cause the ProtocolStack to fail, so the channel constructor will throw an exception
  - start
```
public void start()
           throws java.lang.Exception
```
    从类复制的说明: Protocol
    
    This method is called on a Channel.connect(String). Starts work. Protocols are connected and queues are ready to receive events. Will be called from bottom to top. This call will replace the START and START_OK events.
    
    覆盖:
    
    start 在类中 Protocol
    
    抛出:
    
    java.lang.Exception - Thrown if protocol cannot be started successfully. This will cause the ProtocolStack to fail, so Channel.connect(String) will throw an exception
  - stop
```
public void stop()
```
    从类复制的说明: Protocol
    
    This method is called on a Channel.disconnect(). Stops work (e.g. by closing multicast socket). Will be called from top to bottom. This means that at the time of the method invocation the neighbor protocol below is still working. This method will replace the STOP, STOP_OK, CLEANUP and CLEANUP_OK events. The ProtocolStack guarantees that when this method is called all messages in the down queue will have been flushed
    
    覆盖:
    
    stop 在类中 Protocol
  - up
```
public java.lang.Object up(Event evt)
```
    从类复制的说明: Protocol
    
    An event was received from the layer below. Usually the current layer will want to examine the event type and - depending on its type - perform some computation (e.g. removing headers from a MSG event type, or updating the internal membership list when receiving a VIEW_CHANGE event). Finally the event is either a) discarded, or b) an event is sent down the stack using down_prot.down() or c) the event (or another event) is sent up the stack using up_prot.up().
    
    覆盖:
    
    up 在类中 Protocol
  - handleUpEvent
```
protected void handleUpEvent(STABLE.StableHeader hdr,
                 Address sender,
                 Digest digest)
```
  - up
```
public void up(MessageBatch batch)
```
    从类复制的说明: Protocol
    
    Sends up a multiple messages in a MessageBatch. The sender of the batch is always the same, and so is the destination (null == multicast messages). Messages in a batch can be OOB messages, regular messages, or mixed messages, although the transport itself will create initial MessageBatches that contain only either OOB or regular messages.
    The default processing below sends messages up the stack individually, based on a matching criteria (calling Protocol.accept(Message)), and - if true - calls Protocol.up(Event) for that message and removes the message. If the batch is not empty, it is passed up, or else it is dropped.
    Subclasses should check if there are any messages destined for them (e.g. using MessageBatch.getMatchingMessages(short, boolean)), then possibly remove and process them and finally pass the batch up to the next protocol. Protocols can also modify messages in place, e.g. ENCRYPT could decrypt all encrypted messages in the batch, not remove them, and pass the batch up when done.
    
    覆盖:
    
    up 在类中 Protocol
    
    参数:
    batch - The message batch
  - handleRegularMessage
```
protected void handleRegularMessage(Message msg)
```
  - down
```
public java.lang.Object down(Event evt)
```
    从类复制的说明: Protocol
    
    An event is to be sent down the stack. The layer may want to examine its type and perform some action on it, depending on the event's type. If the event is a message MSG, then the layer may need to add a header to it (or do nothing at all) before sending it down the stack using down_prot.down(). In case of a GET_ADDRESS event (which tries to retrieve the stack's address from one of the bottom layers), the layer may need to send a new response event back up the stack using up_prot.up().
    
    覆盖:
    
    down 在类中 Protocol
  - gc
```
public void gc()
```
  - handleViewChange
```
protected void handleViewChange(View v)
```
  - updateLocalDigest
```
protected void updateLocalDigest(Digest d,
                     Address sender)
```
    Update my own digest from a digest received by somebody else. Returns whether the update was successful. Needs to be called with a lock on digest
  - resetDigest
```
protected void resetDigest()
```
  - addVote
```
protected boolean addVote(int rank)
```
    Adds mbr to votes and returns true if we have all the votes, otherwise false.
    
    参数:
    rank -
  - allVotesReceived
```
protected static boolean allVotesReceived(FixedSizeBitSet votes)
```
    Votes is already locked and guaranteed to be non-null
  - getRank
```
protected static int getRank(Address member,
          View v)
```
  - startStableTask
```
protected void startStableTask()
```
  - stopStableTask
```
protected void stopStableTask()
```
  - startResumeTask
```
protected void startResumeTask(long max_suspend_time)
```
  - stopResumeTask
```
protected void stopResumeTask()
```
  - startStabilityTask
```
protected void startStabilityTask(Digest d,
                      ViewId view_id,
                      long delay)
```
  - stopStabilityTask
```
protected void stopStabilityTask()
```
  - handleStableMessage
```
protected void handleStableMessage(Digest d,
                       Address sender,
                       ViewId view_id)
```
    Digest d contains (a) the highest seqnos deliverable for each sender and (b) the highest seqnos seen for each member. (Difference: with 1,2,4,5, the highest seqno seen is 5, whereas the highest seqno deliverable is 2). The minimum of all highest seqnos deliverable will be taken to send a stability message, which results in garbage collection of messages lower than the ones in the stability vector. The maximum of all seqnos will be taken to trigger possible retransmission of last missing seqno (see DESIGN for details).
  - resetNumBytes
```
protected void resetNumBytes()
```
  - handleStabilityMessage
```
protected void handleStabilityMessage(Digest stable_digest,
                          Address sender,
                          ViewId view_id)
```
  - sendStableMessage
```
protected void sendStableMessage(boolean send_in_background)
```
    Broadcasts a STABLE message of the current digest to all members (or the coordinator only). The message contains the highest seqno delivered and received for all members. The seqnos are retrieved from the NAKACK layer below.
  - marshal
```
public static Buffer marshal(Digest digest)
```
  - readDigest
```
protected Digest readDigest(byte[] buffer,
                int offset,
                int length)
```
  - sendStabilityMessage
```
protected void sendStabilityMessage(Digest tmp,
                        ViewId view_id)
```
    Schedules a stability message to be mcast after a random number of milliseconds (range 1-5 secs). The reason for waiting a random amount of time is that, in the worst case, all members receive a STABLE_GOSSIP message from the last outstanding member at the same time and would therefore mcast the STABILITY message at the same time too. To avoid this, each member waits random N msecs. If, before N elapses, some other member sent the STABILITY message, we just cancel our own message. If, during waiting for N msecs to send STABILITY message S1, another STABILITY message S2 is to be sent, we just discard S2.
    
    参数:
    tmp - A copy of te stability digest, so we don't need to copy it again
  - getDigest
```
protected Digest getDigest()
```
  - printDigest
```
protected java.lang.String printDigest(Digest digest)
```

类 STABLE

嵌套类概要

字段概要

从类继承的字段 com.fr.third.jgroups.stack.Protocol

构造器概要

方法概要

从类继承的方法 com.fr.third.jgroups.stack.Protocol

从类继承的方法 java.lang.Object

字段详细资料

MAX_SUSPEND_TIME

desired_avg_gossip

stability_delay

max_bytes

cap

send_stable_msgs_to_coord_only

num_stable_msgs_sent

num_stable_msgs_received

num_stability_msgs_sent

num_stability_msgs_received

local_addr

view

digest

votes

lock

stability_task_future

stability_lock

stable_task_future

stable_task_lock

timer

num_bytes_received

received

suspended

initialized

resume_task_future

resume_task_mutex

coordinator

构造器详细资料

STABLE

方法详细资料

getDesiredAverageGossip

setDesiredAverageGossip

getMaxBytes

setMaxBytes

getBytes

getStableSent

getStableReceived

getStabilitySent

getStabilityReceived

getStableTaskRunning

resetStats

requiredDownServices

suspend

resume

init

start

stop

up

handleUpEvent

up

handleRegularMessage

down

gc

handleViewChange

updateLocalDigest

resetDigest

addVote

allVotesReceived

getRank

startStableTask

stopStableTask

startResumeTask

stopResumeTask

startStabilityTask

stopStabilityTask

handleStableMessage

resetNumBytes

handleStabilityMessage

sendStableMessage

marshal

readDigest