| <?xml version="1.0" encoding="UTF-8"?> |
| <!-- |
| Copyright 2002-2004 The Apache Software Foundation |
| |
| Licensed under the Apache License, Version 2.0 (the "License"); |
| you may not use this file except in compliance with the License. |
| You may obtain a copy of the License at |
| |
| http://www.apache.org/licenses/LICENSE-2.0 |
| |
| Unless required by applicable law or agreed to in writing, software |
| distributed under the License is distributed on an "AS IS" BASIS, |
| WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| See the License for the specific language governing permissions and |
| limitations under the License. |
| --> |
| |
| <!DOCTYPE article PUBLIC "-//OASIS//DTD Simplified DocBook XML V1.0//EN" |
| "http://www.oasis-open.org/docbook/xml/simple/1.0/sdocbook.dtd"> |
| <article id="ar_Recipes"> |
| <title>ZooKeeper Recipes and Solutions</title> |
| |
| <articleinfo> |
| <legalnotice> |
| <para>Licensed under the Apache License, Version 2.0 (the "License"); |
| you may not use this file except in compliance with the License. You may |
| obtain a copy of the License at <ulink |
| url="http://www.apache.org/licenses/LICENSE-2.0">http://www.apache.org/licenses/LICENSE-2.0</ulink>.</para> |
| |
| <para>Unless required by applicable law or agreed to in writing, |
| software distributed under the License is distributed on an "AS IS" |
| BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or |
| implied. See the License for the specific language governing permissions |
| and limitations under the License.</para> |
| </legalnotice> |
| |
| <abstract> |
| <para>This guide contains pseudocode and guidelines for using Zookeeper to |
| solve common problems in Distributed Application Coordination. It |
| discusses such problems as event handlers, queues, and locks..</para> |
| |
| <para>$Revision: 1.6 $ $Date: 2008/09/19 03:46:18 $</para> |
| </abstract> |
| </articleinfo> |
| |
| <section id="ch_recipes"> |
| <title>A Guide to Creating Higher-level Constructs with ZooKeeper</title> |
| |
| <para>In this article, you'll find guidelines for using |
| ZooKeeper to implement higher order functions. All of them are conventions |
| implemented at the client and do not require special support from |
| ZooKeeper. Hopfully the community will capture these conventions in client-side libraries |
| to ease their use and to encourage standardization.</para> |
| |
| <para>One of the most interesting things about ZooKeeper is that even |
| though ZooKeeper uses <emphasis>asynchronous</emphasis> notifications, you |
| can use it to build <emphasis>synchronous</emphasis> consistency |
| primitives, such as queues and locks. As you will see, this is possible |
| because ZooKeeper imposes an overall order on updates, and has mechanisms |
| to expose this ordering.</para> |
| |
| <para>Note that the recipes below attempt to employ best practices. In |
| particular, they avoid polling, timers or anything else that would result |
| in a "herd effect", causing bursts of traffic and limiting |
| scalability.</para> |
| |
| <para>There are many useful functions that can be imagined that aren't |
| included here - revocable read-write priority locks, as just one example. |
| And some of the constructs mentioned here - locks, in particular - |
| illustrate certain points, even though you may find other constructs, such |
| as event handles or queues, a more practical means of performing the same |
| function. In general, the examples in this section are designed to |
| stimulate thought.</para> |
| |
| |
| <section id="sc_outOfTheBox"> |
| <title>Out of the Box Applications: Name Service, Configuration, Group |
| Membership</title> |
| |
| <para>Name service and configuration are two of the primary applications |
| of ZooKeeper. These two functions are provided directly by the ZooKeeper |
| API.</para> |
| |
| <para>Another function directly provided by ZooKeeper is <emphasis>group |
| membership</emphasis>. The group is represented by a node. Members of the |
| group create ephemeral nodes under the group node. Nodes of the members |
| that fail abnormally will be removed automatically when ZooKeeper detects |
| the failure.</para> |
| </section> |
| |
| <section id="sc_recipes_eventHandles"> |
| <title>Barriers</title> |
| |
| <para>Distributed systems use <emphasis>barriers</emphasis> |
| to block processing of a set of nodes until a condition is met |
| at which time all the nodes are allowed to proceed. Barriers are |
| implemented in ZooKeeper by designating a barrier node. The |
| barrier is in place if the barrier node exists. Here's the |
| pseudo code:</para> |
| |
| <orderedlist> |
| <listitem> |
| <para>Client calls the ZooKeeper API's <emphasis |
| role="bold">exists()</emphasis> function on the barrier node, with |
| <emphasis>watch</emphasis> set to true.</para> |
| </listitem> |
| |
| <listitem> |
| <para>If <emphasis role="bold">exists()</emphasis> returns false, the |
| barrier is gone and the client proceeds</para> |
| </listitem> |
| |
| <listitem> |
| <para>Else, if <emphasis role="bold">exists()</emphasis> returns true, |
| the clients wait for a watch event from ZooKeeper for the barrier |
| node.</para> |
| </listitem> |
| |
| <listitem> |
| <para>When the watch event is triggered, the client reissues the |
| <emphasis role="bold">exists( )</emphasis> call, again waiting until |
| the barrier node is removed.</para> |
| </listitem> |
| </orderedlist> |
| |
| <section id="sc_doubleBarriers"> |
| <title>Double Barriers</title> |
| |
| <para>Double barriers enable clients to synchronize the beginning and |
| the end of a computation. When enough processes have joined the barrier, |
| processes start their computation and leave the barrier once they have |
| finished. This recipe shows how to use a ZooKeeper node as a |
| barrier.</para> |
| |
| <para>The pseudo code in this recipe represents the barrier node as |
| <emphasis>b</emphasis>. Every client process <emphasis>p</emphasis> |
| registers with the barrier node on entry and unregisters when it is |
| ready to leave. A node registers with the barrier node via the <emphasis |
| role="bold">Enter</emphasis> procedure below, it waits until |
| <emphasis>x</emphasis> client process register before proceeding with |
| the computation. (The <emphasis>x</emphasis> here is up to you to |
| determine for your system.)</para> |
| |
| <informaltable colsep="0" frame="none" rowsep="0"> |
| <tgroup cols="2"> |
| <tbody> |
| <row> |
| <entry align="center"><emphasis |
| role="bold">Enter</emphasis></entry> |
| |
| <entry align="center"><emphasis |
| role="bold">Leave</emphasis></entry> |
| </row> |
| |
| <row> |
| <entry align="left"><orderedlist> |
| <listitem> |
| <para>Create a name <emphasis><emphasis>n</emphasis> = |
| <emphasis>b</emphasis>+“/”+<emphasis>p</emphasis></emphasis></para> |
| </listitem> |
| |
| <listitem> |
| <para>Set watch: <emphasis |
| role="bold">exists(<emphasis>b</emphasis> + ‘‘/ready’’, |
| true)</emphasis></para> |
| </listitem> |
| |
| <listitem> |
| <para>Create child: <emphasis role="bold">create( |
| <emphasis>n</emphasis>, EPHEMERAL)</emphasis></para> |
| </listitem> |
| |
| <listitem> |
| <para><emphasis role="bold">L = getChildren(b, |
| false)</emphasis></para> |
| </listitem> |
| |
| <listitem> |
| <para>if fewer children in L than<emphasis> |
| x</emphasis>, wait for watch event</para> |
| </listitem> |
| |
| <listitem> |
| <para>else <emphasis role="bold">create(b + ‘‘/ready’’, |
| REGULAR)</emphasis></para> |
| </listitem> |
| </orderedlist></entry> |
| |
| <entry><orderedlist> |
| <listitem> |
| <para><emphasis role="bold">L = getChildren(b, |
| false)</emphasis></para> |
| </listitem> |
| |
| <listitem> |
| <para>if no children, exit</para> |
| </listitem> |
| |
| <listitem> |
| <para>if <emphasis>p</emphasis> is only process node in |
| L, delete(n) and exit</para> |
| </listitem> |
| |
| <listitem> |
| <para>if <emphasis>p</emphasis> is the lowest process |
| node in L, wait on highest process node in P</para> |
| </listitem> |
| |
| <listitem> |
| <para>else <emphasis |
| role="bold">delete(<emphasis>n</emphasis>) </emphasis>if |
| still exists and wait on lowest process node in L</para> |
| </listitem> |
| |
| <listitem> |
| <para>goto 1</para> |
| </listitem> |
| </orderedlist></entry> |
| </row> |
| </tbody> |
| </tgroup> |
| </informaltable> |
| <para>On entering, all processes watch on a ready node and |
| create an ephemeral node as a child of the barrier node. Each process |
| but the last enters the barrier and waits for the ready node to appear |
| at line 5. The process that creates the xth node, the last process, will |
| see x nodes in the list of children and create the ready node, waking up |
| the other processes. Note that waiting processes wake up only when it is |
| time to exit, so waiting is efficient. |
| </para> |
| |
| <para>On exit, you can't use a flag such as <emphasis>ready</emphasis> |
| because you are watching for process nodes to go away. By using |
| ephemeral nodes, processes that fail after the barrier has been entered |
| do not prevent correct processes from finishing. When processes are |
| ready to leave, they need to delete their process nodes and wait for all |
| other processes to do the same.</para> |
| |
| <para>Processes exit when there are no process nodes left as children of |
| <emphasis>b</emphasis>. However, as an efficiency, you can use the |
| lowest process node as the ready flag. All other processes that are |
| ready to exit watch for the lowest existing process node to go away, and |
| the owner of the lowest process watches for any other process node |
| (picking the highest for simplicity) to go away. This means that only a |
| single process wakes up on each node deletion except for the last node, |
| which wakes up everyone when it is removed.</para> |
| </section> |
| </section> |
| |
| <section id="sc_recipes_Queues"> |
| <title>Queues</title> |
| |
| <para>Distributed queues are a common data structure. To implement a |
| distributed queue in ZooKeeper, first designate a znode to hold the queue, |
| the queue node. The distributed clients put something into the queue by |
| calling create() with a pathname ending in "queue-", with the |
| <emphasis>sequence</emphasis> and <emphasis>ephemeral</emphasis> flags in |
| the create() call set to true. Because the <emphasis>sequence</emphasis> |
| flag is set, the new pathnames will have the form |
| _path-to-queue-node_/queue-X, where X is a monotonic increasing number. A |
| client that wants to be removed from the queue calls ZooKeeper's <emphasis |
| role="bold">getChildren( )</emphasis> function, with |
| <emphasis>watch</emphasis> set to true on the queue node, and begins |
| processing nodes with the lowest number. The client does not need to issue |
| another <emphasis role="bold">getChildren( )</emphasis> until it exhausts |
| the list obtained from the first <emphasis role="bold">getChildren( |
| )</emphasis> call. If there are are no children in the queue node, the |
| reader waits for a watch notification to check the queue again.</para> |
| |
| <note> |
| <para>There now exists a Queue implementation in ZooKeeper |
| recipes directory. This is distributed with the release -- |
| src/recipes/queue directory of the release artifact. |
| </para> |
| </note> |
| |
| <section id="sc_recipes_priorityQueues"> |
| <title>Priority Queues</title> |
| |
| <para>To implement a priority queue, you need only make two simple |
| changes to the generic <ulink url="#sc_recipes_Queues">queue |
| recipe</ulink> . First, to add to a queue, the pathname ends with |
| "queue-YY" where YY is the priority of the element with lower numbers |
| representing higher priority (just like UNIX). Second, when removing |
| from the queue, a client uses an up-to-date children list meaning that |
| the client will invalidate previously obtained children lists if a watch |
| notification triggers for the queue node.</para> |
| </section> |
| </section> |
| |
| <section id="sc_recipes_Locks"> |
| <title>Locks</title> |
| |
| <para>Fully distributed locks that are globally synchronous, meaning at |
| any snapshot in time no two clients think they hold the same lock. These |
| can be implemented using ZooKeeeper. As with priority queues, first define |
| a lock node.</para> |
| |
| <note> |
| <para>There now exists a Lock implementation in ZooKeeper |
| recipes directory. This is distributed with the release -- |
| src/recipes/lock directory of the release artifact. |
| </para> |
| </note> |
| |
| <para>Clients wishing to obtain a lock do the following:</para> |
| |
| <orderedlist> |
| <listitem> |
| <para>Call <emphasis role="bold">create( )</emphasis> with a pathname |
| of "_locknode_/lock-" and the <emphasis>sequence</emphasis> and |
| <emphasis>ephemeral</emphasis> flags set.</para> |
| </listitem> |
| |
| <listitem> |
| <para>Call <emphasis role="bold">getChildren( )</emphasis> on the lock |
| node <emphasis>without</emphasis> setting the watch flag (this is |
| important to avoid the herd effect).</para> |
| </listitem> |
| |
| <listitem> |
| <para>If the pathname created in step <emphasis |
| role="bold">1</emphasis> has the lowest sequence number suffix, the |
| client has the lock and the client exits the protocol.</para> |
| </listitem> |
| |
| <listitem> |
| <para>The client calls <emphasis role="bold">exists( )</emphasis> with |
| the watch flag set on the path in the lock directory with the next |
| lowest sequence number.</para> |
| </listitem> |
| |
| <listitem> |
| <para>if <emphasis role="bold">exists( )</emphasis> returns false, go |
| to step <emphasis role="bold">2</emphasis>. Otherwise, wait for a |
| notification for the pathname from the previous step before going to |
| step <emphasis role="bold">2</emphasis>.</para> |
| </listitem> |
| </orderedlist> |
| |
| <para>The unlock protocol is very simple: clients wishing to release a |
| lock simply delete the node they created in step 1.</para> |
| |
| <para>Here are a few things to notice:</para> |
| |
| <itemizedlist> |
| <listitem> |
| <para>The removal of a node will only cause one client to wake up |
| since each node is watched by exactly one client. In this way, you |
| avoid the herd effect.</para> |
| </listitem> |
| </itemizedlist> |
| |
| <itemizedlist> |
| <listitem> |
| <para>There is no polling or timeouts.</para> |
| </listitem> |
| </itemizedlist> |
| |
| <itemizedlist> |
| <listitem> |
| <para>Because of the way you implement locking, it is easy to see the |
| amount of lock contention, break locks, debug locking problems, |
| etc.</para> |
| </listitem> |
| </itemizedlist> |
| |
| <section> |
| <title>Shared Locks</title> |
| |
| <para>You can implement shared locks by with a few changes to the lock |
| protocol:</para> |
| |
| <informaltable colsep="0" frame="none" rowsep="0"> |
| <tgroup cols="2"> |
| <tbody> |
| <row> |
| <entry align="center"><emphasis role="bold">Obtaining a read |
| lock:</emphasis></entry> |
| |
| <entry align="center"><emphasis role="bold">Obtaining a write |
| lock:</emphasis></entry> |
| </row> |
| |
| <row> |
| <entry align="left"><orderedlist> |
| <listitem> |
| <para>Call <emphasis role="bold">create( )</emphasis> to |
| create a node with pathname |
| "<filename>_locknode_/read-</filename>". This is the |
| lock node use later in the protocol. Make sure to set both |
| the <emphasis>sequence</emphasis> and |
| <emphasis>ephemeral</emphasis> flags.</para> |
| </listitem> |
| |
| <listitem> |
| <para>Call <emphasis role="bold">getChildren( )</emphasis> |
| on the lock node <emphasis>without</emphasis> setting the |
| <emphasis>watch</emphasis> flag - this is important, as it |
| avoids the herd effect.</para> |
| </listitem> |
| |
| <listitem> |
| <para>If there are no children with a pathname starting |
| with "<filename>write-</filename>" and having a lower |
| sequence number than the node created in step <emphasis |
| role="bold">1</emphasis>, the client has the lock and can |
| exit the protocol. </para> |
| </listitem> |
| |
| <listitem> |
| <para>Otherwise, call <emphasis role="bold">exists( |
| )</emphasis>, with <emphasis>watch</emphasis> flag, set on |
| the node in lock directory with pathname staring with |
| "<filename>write-</filename>" having the next lowest |
| sequence number.</para> |
| </listitem> |
| |
| <listitem> |
| <para>If <emphasis role="bold">exists( )</emphasis> |
| returns <emphasis>false</emphasis>, goto step <emphasis |
| role="bold">2</emphasis>.</para> |
| </listitem> |
| |
| <listitem> |
| <para>Otherwise, wait for a notification for the pathname |
| from the previous step before going to step <emphasis |
| role="bold">2</emphasis></para> |
| </listitem> |
| </orderedlist></entry> |
| |
| <entry><orderedlist> |
| <listitem> |
| <para>Call <emphasis role="bold">create( )</emphasis> to |
| create a node with pathname |
| "<filename>_locknode_/write-</filename>". This is the |
| lock node spoken of later in the protocol. Make sure to |
| set both <emphasis>sequence</emphasis> and |
| <emphasis>ephemeral</emphasis> flags.</para> |
| </listitem> |
| |
| <listitem> |
| <para>Call <emphasis role="bold">getChildren( ) |
| </emphasis> on the lock node <emphasis>without</emphasis> |
| setting the <emphasis>watch</emphasis> flag - this is |
| important, as it avoids the herd effect.</para> |
| </listitem> |
| |
| <listitem> |
| <para>If there are no children with a lower sequence |
| number than the node created in step <emphasis |
| role="bold">1</emphasis>, the client has the lock and the |
| client exits the protocol.</para> |
| </listitem> |
| |
| <listitem> |
| <para>Call <emphasis role="bold">exists( ),</emphasis> |
| with <emphasis>watch</emphasis> flag set, on the node with |
| the pathname that has the next lowest sequence |
| number.</para> |
| </listitem> |
| |
| <listitem> |
| <para>If <emphasis role="bold">exists( )</emphasis> |
| returns <emphasis>false</emphasis>, goto step <emphasis |
| role="bold">2</emphasis>. Otherwise, wait for a |
| notification for the pathname from the previous step |
| before going to step <emphasis |
| role="bold">2</emphasis>.</para> |
| </listitem> |
| </orderedlist></entry> |
| </row> |
| </tbody> |
| </tgroup> |
| </informaltable> |
| |
| <note> |
| <para>It might appear that this recipe creates a herd effect: |
| when there is a large group of clients waiting for a read |
| lock, and all getting notified more or less simultaneously |
| when the "<filename>write-</filename>" node with the lowest |
| sequence number is deleted. In fact. that's valid behavior: |
| as all those waiting reader clients should be released since |
| they have the lock. The herd effect refers to releasing a |
| "herd" when in fact only a single or a small number of |
| machines can proceed. |
| </para> |
| </note> |
| </section> |
| |
| <section id="sc_recoverableSharedLocks"> |
| <title>Recoverable Shared Locks</title> |
| |
| <para>With minor modifications to the Shared Lock protocol, you make |
| shared locks revocable by modifying the shared lock protocol:</para> |
| |
| <para>In step <emphasis role="bold">1</emphasis>, of both obtain reader |
| and writer lock protocols, call <emphasis role="bold">getData( |
| )</emphasis> with <emphasis>watch</emphasis> set, immediately after the |
| call to <emphasis role="bold">create( )</emphasis>. If the client |
| subsequently receives notification for the node it created in step |
| <emphasis role="bold">1</emphasis>, it does another <emphasis |
| role="bold">getData( )</emphasis> on that node, with |
| <emphasis>watch</emphasis> set and looks for the string "unlock", which |
| signals to the client that it must release the lock. This is because, |
| according to this shared lock protocol, you can request the client with |
| the lock give up the lock by calling <emphasis role="bold">setData() |
| </emphasis> on the lock node, writing "unlock" to that node.</para> |
| |
| <para>Note that this protocol requires the lock holder to consent to |
| releasing the lock. Such consent is important, especially if the lock |
| holder needs to do some processing before releasing the lock. Of course |
| you can always implement <emphasis>Revocable Shared Locks with Freaking |
| Laser Beams</emphasis> by stipulating in your protocol that the revoker |
| is allowed to delete the lock node if after some length of time the lock |
| isn't deleted by the lock holder.</para> |
| </section> |
| </section> |
| |
| <section id="sc_recipes_twoPhasedCommit"> |
| <title>Two-phased Commit</title> |
| |
| <para>A two-phase commit protocol is an algorithm that lets all clients in |
| a distributed system agree either to commit a transaction or abort.</para> |
| |
| <para>In ZooKeeper, you can implement a two-phased commit by having a |
| coordinator create a transaction node, say "/app/Tx", and one child node |
| per participating site, say "/app/Tx/s_i". When coordinator creates the |
| child node, it leaves the content undefined. Once each site involved in |
| the transaction receives the transaction from the coordinator, the site |
| reads each child node and sets a watch. Each site then processes the query |
| and votes "commit" or "abort" by writing to its respective node. Once the |
| write completes, the other sites are notified, and as soon as all sites |
| have all votes, they can decide either "abort" or "commit". Note that a |
| node can decide "abort" earlier if some site votes for "abort".</para> |
| |
| <para>An interesting aspect of this implementation is that the only role |
| of the coordinator is to decide upon the group of sites, to create the |
| ZooKeeper nodes, and to propagate the transaction to the corresponding |
| sites. In fact, even propagating the transaction can be done through |
| ZooKeeper by writing it in the transaction node.</para> |
| |
| <para>There are two important drawbacks of the approach described above. |
| One is the message complexity, which is O(n²). The second is the |
| impossibility of detecting failures of sites through ephemeral nodes. To |
| detect the failure of a site using ephemeral nodes, it is necessary that |
| the site create the node.</para> |
| |
| <para>To solve the first problem, you can have only the coordinator |
| notified of changes to the transaction nodes, and then notify the sites |
| once coordinator reaches a decision. Note that this approach is scalable, |
| but it's is slower too, as it requires all communication to go through the |
| coordinator.</para> |
| |
| <para>To address the second problem, you can have the coordinator |
| propagate the transaction to the sites, and have each site creating its |
| own ephemeral node.</para> |
| </section> |
| |
| <section id="sc_leaderElection"> |
| <title>Leader Election</title> |
| |
| <para>A simple way of doing leader election with ZooKeeper is to use the |
| <emphasis role="bold">SEQUENCE|EPHEMERAL</emphasis> flags when creating |
| znodes that represent "proposals" of clients. The idea is to have a znode, |
| say "/election", such that each znode creates a child znode "/election/n_" |
| with both flags SEQUENCE|EPHEMERAL. With the sequence flag, ZooKeeper |
| automatically appends a sequence number that is greater that any one |
| previously appended to a child of "/election". The process that created |
| the znode with the smallest appended sequence number is the leader. |
| </para> |
| |
| <para>That's not all, though. It is important to watch for failures of the |
| leader, so that a new client arises as the new leader in the case the |
| current leader fails. A trivial solution is to have all application |
| processes watching upon the current smallest znode, and checking if they |
| are the new leader when the smallest znode goes away (note that the |
| smallest znode will go away if the leader fails because the node is |
| ephemeral). But this causes a herd effect: upon of failure of the current |
| leader, all other processes receive a notification, and execute |
| getChildren on "/election" to obtain the current list of children of |
| "/election". If the number of clients is large, it causes a spike on the |
| number of operations that ZooKeeper servers have to process. To avoid the |
| herd effect, it is sufficient to watch for the next znode down on the |
| sequence of znodes. If a client receives a notification that the znode it |
| is watching is gone, then it becomes the new leader in the case that there |
| is no smaller znode. Note that this avoids the herd effect by not having |
| all clients watching the same znode. </para> |
| |
| <para>Here's the pseudo code:</para> |
| |
| <para>Let ELECTION be a path of choice of the application. To volunteer to |
| be a leader: </para> |
| |
| <orderedlist> |
| <listitem> |
| <para>Create znode z with path "ELECTION/n_" with both SEQUENCE and |
| EPHEMERAL flags;</para> |
| </listitem> |
| |
| <listitem> |
| <para>Let C be the children of "ELECTION", and i be the sequence |
| number of z;</para> |
| </listitem> |
| |
| <listitem> |
| <para>Watch for changes on "ELECTION/n_j", where j is the smallest |
| sequence number such that j < i and n_j is a znode in C;</para> |
| </listitem> |
| </orderedlist> |
| |
| <para>Upon receiving a notification of znode deletion: </para> |
| |
| <orderedlist> |
| <listitem> |
| <para>Let C be the new set of children of ELECTION; </para> |
| </listitem> |
| |
| <listitem> |
| <para>If z is the smallest node in C, then execute leader |
| procedure;</para> |
| </listitem> |
| |
| <listitem> |
| <para>Otherwise, watch for changes on "ELECTION/n_j", where j is the |
| smallest sequence number such that j < i and n_j is a znode in C; |
| </para> |
| </listitem> |
| </orderedlist> |
| |
| <para>Note that the znode having no preceding znode on the list of |
| children does not imply that the creator of this znode is aware that it is |
| the current leader. Applications may consider creating a separate to znode |
| to acknowledge that the leader has executed the leader procedure. </para> |
| </section> |
| </section> |
| </article> |