Hello, One more thing, if i leave readquorum to 1 and write to 2 and add more nodes, like 3 more, how will it work, i mean if i then resize quorumm size. it will only replicate on newly added databases or?
On Thursday, July 16, 2015 at 4:20:20 PM UTC+2, Matija Žeželj wrote: > > Hello, > > With readquorum set to 1 and write to 2 it works, but it cannot guarantee > data consistency. or am i wrong? > Strange thing is on customers box with exactly the same setitngs (same > orientdb version) it works and on ours it always fails. > I tested it on debian wheezy and orache java 8 and later on on openjdk 7 > and same errors. > > Have any idea what might be the issue here? > > Thanks! > > On Wednesday, July 15, 2015 at 11:36:19 PM UTC+2, l.garulli wrote: >> >> Hi Matija, >> Could you please try with readQuorum=1 and writeQuorum=2? >> >> Best Regards, >> >> Founder & CEO >> OrientDB <http://orientdb.com/> >> >> >> On 15 July 2015 at 20:47, Matija Žeželj <[email protected]> wrote: >> >>> Hello, >>> >>> i'm using 2.1 rc5 and i have 3 nodes in cluster. >>> setup: >>> in hazelcast i use tcp-ip to connect to nodes, on each node in >>> hazelcats.xml i have 2 other nodes listed >>> in default-distributed-db-config.json i have "readQuorum": 2 , other >>> values are set to default ones >>> and in orientdb-server.config i have set hazelcast to be enabled and >>> node name(node1-3) >>> my couse of action: >>> started server1, added db with create database >>> plocal:../databases/SOMEDB USER PASS plocal graph >>> stopped node 1, sttarted node1, started node2, it synced, started node3 >>> , it synced. >>> when i try and open webinterface and login as root to db and go to >>> security i get : >>> 2015-07-15 10:23:14:153 WARNI ->MYIP: Command not found: >>> GET.sqlProfiler/Events [ONetworkProtocolHttpDb] >>> 2015-07-15 10:23:14:192 SEVER Internal server error: >>> com.orientechnologies.orient.core.exception.OQueryParsingException: >>> Error on parsing query at position #8: Error on parsing command at position >>> #0: Error on parsing query >>> Query: _studio WHERE user.name = '{{username}}' AND type = >>> 'GraphConfig' >>> -------------^ >>> --> >>> com.orientechnologies.orient.core.exception.OCommandExecutionException: >>> Class '_STUDIO' was not found in current database [ONetworkProtocolHttpDb] >>> 2015-07-15 10:23:28:238 WARNI [node1] timeout (9003ms) on waiting for >>> synchronous responses from nodes=[node2, node3, node1] responsesSoFar=[] >>> request=id=3 from=node1 task=record_read [OHazelcastDistributedDatabase] >>> 2015-07-15 10:23:28:240 WARNI [node1] no response received from local >>> node about request id=3 from=node1 task=record_read >>> [ODistributedResponseManager] >>> 2015-07-15 10:23:28:241 WARNI [node1] detected 3 node(s) in timeout or >>> in conflict and quorum (2) has not been reached, rolling back changes for >>> request (id=3 from=node1 task=record_read) [ODistributedResponseManager] >>> 2015-07-15 10:23:28:241 WARNI [node1] Quorum 2 not reached for request >>> (id=3 from=node1 task=record_read). Elapsed=9017ms No server in conflict. >>> Received: {node2=waiting-for-response, node3=waiting-for-response, >>> node1=waiting-for-response} [ODistributedResponseManager] >>> 2015-07-15 10:23:28:243 SEVER Internal server error: >>> com.orientechnologies.orient.server.distributed.ODistributedException: >>> Quorum 2 not reached for request (id=3 from=node1 task=record_read). >>> Elapsed=9017ms No server in conflict. Received: >>> {node2=waiting-for-response, node3=waiting-for-response, >>> node1=waiting-for-response} [ONetworkProtocolHttpDb] >>> 2015-07-15 10:23:28:244 WARNI [node1] timeout (9003ms) on waiting for >>> synchronous responses from nodes=[node2, node3, node1] responsesSoFar=[] >>> request=id=4 from=node1 task=record_read [OHazelcastDistributedDatabase] >>> 2015-07-15 10:23:28:245 WARNI [node1] no response received from local >>> node about request id=4 from=node1 task=record_read >>> [ODistributedResponseManager] >>> 2015-07-15 10:23:28:245 WARNI [node1] detected 3 node(s) in timeout or >>> in conflict and quorum (2) has not been reached, rolling back changes for >>> request (id=4 from=node1 task=record_read) [ODistributedResponseManager] >>> 2015-07-15 10:23:28:246 WARNI [node1] Quorum 2 not reached for request >>> (id=4 from=node1 task=record_read). Elapsed=9021ms No server in conflict. >>> Received: {node2=waiting-for-response, node3=waiting-for-response, >>> node1=waiting-for-response} [ODistributedResponseM >>> >>> And it loops on all nodes. when i stop and restart all nodes dbs sync >>> and it works again until i start web interface and go to security tab. I >>> tried adding users there but as you can see ^^ no luck :/ >>> >>> Any ideas? >>> maybe start server on node1 before setting distributed to enabled, add >>> db, go to web interface add users do what i need. stop node1, start it >>> distribuded, start node 2, wait , start node 3? >>> >>> I never worked with orientdb, i just need to set it up for customer. >>> >>> -- >>> >>> --- >>> You received this message because you are subscribed to the Google >>> Groups "OrientDB" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- --- You received this message because you are subscribed to the Google Groups "OrientDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
