[ https://issues.apache.org/jira/browse/IGNITE-2355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15554712#comment-15554712 ]
Taras Ledkov edited comment on IGNITE-2355 at 10/7/16 12:01 PM: ---------------------------------------------------------------- Probably configuration properties, according with HDFS HA & Cloudera map-reduce HA: {code} <property> <name>mapred.job.tracker</name> <value>ignite0</value> <!-- Possible values are <host:port> string or logical name of Ignite cluster --> </property> <property> <name>ignite.servers.ignite0</name> <value>srv0,srv1</value> </property> <property> <name>ignite.server-addr.ignite0.srv0</name> <value><host0>:<port0></value> </property> <property> <name>ignite.server-addr.ignite0.srv1</name> <value><host1>:<port1></value> </property> ... {code} was (Author: tledkov-gridgain): Probably configuration properties, according with HDFS HA & Cloudera map-reduce HA: {code} <property> <name>mapred.job.tracker</name> <value>ignite0</value> <!-- Possible values are <host:port> string or list or logical name of Ignite cluster --> </property> <property> <name>ignite.servers.ignite0</name> <value>srv0,srv1</value> </property> <property> <name>ignite.server-addr.ignite0.srv0</name> <value><host0>:<port0></value> </property> <property> <name>ignite.server-addr.ignite0.srv1</name> <value><host1>:<port1></value> </property> ... {code} > Hadoop client should be able to failover in case of server crash. > ----------------------------------------------------------------- > > Key: IGNITE-2355 > URL: https://issues.apache.org/jira/browse/IGNITE-2355 > Project: Ignite > Issue Type: Sub-task > Components: hadoop > Affects Versions: ignite-1.4 > Reporter: Vladimir Ozerov > Assignee: Taras Ledkov > Priority: Critical > Labels: roadmap > Fix For: 2.0 > > > Currently we explicitly specify a single IP address of Ignite server for map > reduce. > If it goes down, no jobs can be submitted anymore. > Looks like we should give user ability to specify multiple addresses, and > failover between them. Our thin client (which underlies Hadoop client) is > already able to accept multiple addresses. > Look at Hadoop HA first to get an ideas on how to configure multiple > addresses. -- This message was sent by Atlassian JIRA (v6.3.4#6332)