| layout | displayTitle | title | description | license |
|---|---|---|---|---|
page |
Uniffle Coordinator Guide |
Uniffle Coordinator Guide |
Uniffle Coordinator Guide |
Licensed to the Apache Software Foundation (ASF) under one or more
contributor license agreements. See the NOTICE file distributed with
this work for additional information regarding copyright ownership.
The ASF licenses this file to You 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.
|
Uniffle is a unified remote shuffle service for compute engines, the role of coordinator is responsibility for collecting status of shuffle server and doing the assignment for the job.
This document will introduce how to deploy Uniffle coordinators.
-
unzip package to RSS_HOME
-
update RSS_HOME/bin/rss-env.sh, eg,
JAVA_HOME=<java_home> HADOOP_HOME=<hadoop home> XMX_SIZE="16g" -
update RSS_HOME/conf/coordinator.conf, eg,
rss.rpc.server.port 19999 rss.jetty.http.port 19998 rss.coordinator.server.heartbeat.timeout 30000 rss.coordinator.app.expired 60000 rss.coordinator.shuffle.nodes.max 5 # enable dynamicClientConf, and coordinator will be responsible for most of client conf rss.coordinator.dynamicClientConf.enabled true # config the path of client conf rss.coordinator.dynamicClientConf.path <RSS_HOME>/conf/dynamic_client.conf # config the path of excluded shuffle server rss.coordinator.exclude.nodes.file.path <RSS_HOME>/conf/exclude_nodes -
update <RSS_HOME>/conf/dynamic_client.conf, rss client will get default conf from coordinator eg,
# MEMORY_LOCALFILE_HDFS is recommandation for production environment rss.storage.type MEMORY_LOCALFILE_HDFS # multiple remote storages are supported, and client will get assignment from coordinator rss.coordinator.remote.storage.path hdfs://cluster1/path,hdfs://cluster2/path rss.writer.require.memory.retryMax 1200 rss.client.retry.max 100 rss.writer.send.check.timeout 600000 rss.client.read.buffer.size 14m -
update <RSS_HOME>/conf/exclude_nodes, coordinator will update excluded node by this file eg,
# shuffleServer's ip and port, connected with "-" 110.23.15.36-19999 110.23.15.35-19996 -
start Coordinator
bash RSS_HOME/bin/start-coordnator.sh
| Property Name | Default | Description |
|---|---|---|
| rss.coordinator.server.heartbeat.timeout | 30000 | Timeout if can't get heartbeat from shuffle server |
| rss.coordinator.server.periodic.output.interval.times | 30 | The periodic interval times of output alive nodes. The interval sec can be calculated by (rss.coordinator.server.heartbeat.timeout/3 * rss.coordinator.server.periodic.output.interval.times). Default output interval is 5min. |
| rss.coordinator.assignment.strategy | PARTITION_BALANCE | Strategy for assigning shuffle server, PARTITION_BALANCE should be used for workload balance |
| rss.coordinator.app.expired | 60000 | Application expired time (ms), the heartbeat interval should be less than it |
| rss.coordinator.shuffle.nodes.max | 9 | The max number of shuffle server when do the assignment |
| rss.coordinator.dynamicClientConf.path | - | The path of configuration file which have default conf for rss client |
| rss.coordinator.exclude.nodes.file.path | - | The path of configuration file which have exclude nodes |
| rss.coordinator.exclude.nodes.check.interval.ms | 60000 | Update interval (ms) for exclude nodes |
| rss.coordinator.access.checkers | org.apache.uniffle.coordinator.access.checker.AccessClusterLoadChecker | The access checkers will be used when the spark client use the DelegationShuffleManager, which will decide whether to use rss according to the result of the specified access checkers |
| rss.coordinator.access.loadChecker.memory.percentage | 15.0 | The minimal percentage of available memory percentage of a server |
| rss.coordinator.dynamicClientConf.enabled | false | whether to enable dynamic client conf, which will be fetched by spark client |
| rss.coordinator.dynamicClientConf.path | - | The dynamic client conf of this cluster and can be stored in HADOOP FS or local |
| rss.coordinator.dynamicClientConf.updateIntervalSec | 120 | The dynamic client conf update interval in seconds |
| rss.coordinator.remote.storage.cluster.conf | - | Remote Storage Cluster related conf with format $clusterId,$key=$value, separated by ';' |
| rss.rpc.server.port | - | RPC port for coordinator |
| rss.jetty.http.port | - | Http port for coordinator |
| rss.coordinator.remote.storage.select.strategy | APP_BALANCE | Strategy for selecting the remote path |
| rss.coordinator.remote.storage.io.sample.schedule.time | 60000 | The time of scheduling the read and write time of the paths to obtain different HADOOP FS |
| rss.coordinator.remote.storage.io.sample.file.size | 204800000 | The size of the file that the scheduled thread reads and writes |
| rss.coordinator.remote.storage.io.sample.access.times | 3 | The number of times to read and write HADOOP FS files |
| rss.coordinator.startup-silent-period.enabled | false | Enable the startup-silent-period to reject the assignment requests for avoiding partial assignments. To avoid service interruption, this mechanism is disabled by default. Especially it's recommended to use in coordinator HA mode when restarting single coordinator. |
| rss.coordinator.startup-silent-period.duration | 20000 | The waiting duration(ms) when conf of rss.coordinator.startup-silent-period.enabled is enabled. |
| rss.coordinator.select.partition.strategy | ROUND | There are two strategies for selecting partitions: ROUND and CONTINUOUS. ROUND will poll to allocate partitions to ShuffleServer, and CONTINUOUS will try to allocate consecutive partitions to ShuffleServer, this feature can improve performance in AQE scenarios. |
| rss.metrics.reporter.class | - | The class of metrics reporter. |
| rss.reconfigure.interval.sec | 5 | Reconfigure check interval. |
| Property Name | Default | Description |
|---|---|---|
| rss.coordinator.access.loadChecker.serverNum.threshold | - | The minimal required number of healthy shuffle servers when being accessed by client. And when not specified, it will use the required shuffle-server number from client as the checking condition. If there is no client shuffle-server number specified, the coordinator conf of rss.coordinator.shuffle.nodes.max will be adopted |
AccessCandidatesChecker is one of the built-in access checker, which will allow user to define the candidates list to use rss.
| Property Name | Default | Description |
|---|---|---|
| rss.coordinator.access.candidates.updateIntervalSec | 120 | Accessed candidates update interval in seconds, which is only valid when AccessCandidatesChecker is enabled. |
| rss.coordinator.access.candidates.path | - | Accessed candidates file path, the file can be stored on HADOOP FS |
AccessQuotaChecker is a checker when the number of concurrent tasks submitted by users increases sharply, some important apps may be affected. Therefore, we restrict users to submit to the uniffle cluster, and rejected apps will be submitted to ESS.
| Property Name | Default | Description |
|---|---|---|
| rss.coordinator.quota.update.interval | 60000 | Update interval for the default number of submitted apps per user. |
| rss.coordinator.quota.default.path | - | A configuration file for the number of apps for a user-defined user. |
| rss.coordinator.quota.default.app.num | 5 | Default number of apps at user level. |
GET /api/server/nodes/{id}
name type data type description id required string shuffle server id, eg:127.0.0.1-19999
curl -X GET http://localhost:19998/api/server/nodes/127.0.0.1-19999
GET /api/server/nodes
name type data type description status optional string Shuffle server status, eg:ACTIVE, DECOMMISSIONING, DECOMMISSIONED
curl -X GET http://localhost:19998/api/server/nodes curl -X GET http://localhost:19998/api/server/nodes?status=ACTIVE
POST /api/server/decommission
name type data type description serverIds required array Shuffle server array, eg:["127.0.0.1-19999"]
curl -X POST -H "Content-Type: application/json" http://localhost:19998/api/server/decommission -d '{"serverIds:": ["127.0.0.1:19999"]}'
POST /api/server/{id}/decommission
name type data type description id required string Shuffle server id, eg:127.0.0.1-19999
curl -X POST -H "Content-Type: application/json" http://localhost:19998/api/server/127.0.0.1-19999/decommission
POST /api/server/cancelDecommission
name type data type description serverIds required array Shuffle server array, eg:["127.0.0.1:19999"]
curl -X POST -H "Content-Type: application/json" http://localhost:19998/api/server/cancelDecommission -d '{"serverIds:": ["127.0.0.1:19999"]}'