Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: <derby-dev@db.apache.org>
Message-ID: <1667469987.1204014831206.JavaMail.jira@brutus>
Date: Tue, 26 Feb 2008 00:33:51 -0800 (PST)
From: =?utf-8?Q?J=C3=B8rgen_L=C3=B8land_=28JIRA=29?= <jira@apache.org>
To: derby-dev@db.apache.org
Subject: [jira] Commented: (DERBY-3463) Slave db is still in replication
 slave mode after failover if there was replication load before failover.
In-Reply-To: <1688533851.1203971932001.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/DERBY-3463?page=3Dcom.atlassian=
.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D1257=
2383#action_12572383 ]=20

J=C3=B8rgen L=C3=B8land commented on DERBY-3463:
--------------------------------------

Hi Ole,

I tried to reproduce the first problem (slave still in slave mode after fai=
lover), but running a simple test client (insert/update/delete) works fine =
without reproducing the bug. The last line of the reported exception looks =
interesting:=20

"Caused by: ERROR XSDFI: Error encountered while trying to write data to di=
sk during database recovery. Check that the database disk is not full. If i=
t is then delete unnecessary files, and retry connecting to the database. I=
t is also possible that the file system is read only, or the disk has faile=
d, or some other problem with the media. System encountered error while pro=
cessing page Page(2,Container(0, 336)). "

Can you verify that the disk is not full and is working, e.g. by rerunning =
the test on another machine or disk?=20

If the disk is not the problem:=20
I thought all operation types were working with replication, but this may n=
ot be right. Hence, it could be the case that your workload contains an ope=
ration that is not handled correctly. Can you provide some more information=
 on the workload (or even better - check if there are any specific operatio=
ns that causes the problem)?

> Slave db is still in replication slave mode after failover if there was r=
eplication load before failover.
> -------------------------------------------------------------------------=
--------------------------------
>
>                 Key: DERBY-3463
>                 URL: https://issues.apache.org/jira/browse/DERBY-3463
>             Project: Derby
>          Issue Type: Bug
>          Components: Replication
>         Environment: Trunk, svn r630806
>            Reporter: Ole Solberg
>            Assignee: J=C3=B8rgen L=C3=B8land
>             Fix For: 10.4.0.0
>
>         Attachments: NOreplicationLoad.tar, replicationLoad.tar
>
>
> Slave db is still in replication slave mode after failover if there was r=
eplication load before failover.
> I.e. connect to slave after failover is refused. Case A) below.
> (Connect to master after failover is OK.)
> If however there is no replication load on master db before failover conn=
ection to slave succeeds,
> but connection to master fails(hang).
> Trunk svn r630806
> A) Replication load.
> -------------------
> Master and slave servers started.
> Master: create master db.
> Master: freeze db
> Copy master db to slave.
> Master: unfreeze db.
> Slave:
> jdbc:derby://localhost:4527//home/os136789/Replication/testing/db_slave/w=
ombat;startSlave=3Dtrue;slaveHost=3Dlocalhost;slavePort=3D6666
> Master:
> jdbc:derby://localhost:1527//home/os136789/Replication/testing/db_master/=
wombat;startMaster=3Dtrue;slaveHost=3Dlocalhost;slavePort=3D6666
> Slave:
> SQLException: -1 XRE08 DERBY SQL error: SQLCODE: -1, SQLSTATE: XRE08, SQL=
ERRMC: Replication slave mode started successfully for database '/home/os13=
6789/Replication/testing/db_slave/wombat'. Connection refused because the d=
atabase is in replication slave mode.
> Running tests against master.
> Master:
> jdbc:derby://localhost:1527//home/os136789/Replication/testing/db_master/=
wombat;failover=3Dtrue
> SQLException: -1 XRE20 DERBY SQL error: SQLCODE: -1, SQLSTATE: XRE20, SQL=
ERRMC: Failover performed successfully for database '/home/os136789/Replica=
tion/testing/db_master/wombat', the database has been shutdown.
> Slave:
> /usr/local/java/jdk1.6/jre/lib/../bin/java -Dij.driver=3Dorg.apache.derby=
.jdbc.ClientDriver -Dij.connection.slave=3Djdbc:derby://localhost:4527//hom=
e/os136789/Replication/testing/db_slave/wombat -classpath /home/os136789/Ap=
ache/myDerbySandbox/trunk/jars/insane/derbyclient.jar:/home/os136789/Apache=
/myDerbySandbox/trunk/jars/insane/derbytools.jar org.apache.derby.tools.ij =
/home/os136789/Replication/testing/verify_ReplicationTestRun.sql
> ERROR 08004: DERBY SQL error: SQLCODE: -1, SQLSTATE: 08004, SQLERRMC: Con=
nection refused to database '/home/os136789/Replication/testing/db_slave/wo=
mbat' because it is in replication slave mode.
> Master:
> /usr/local/java/jdk1.6/jre/lib/../bin/java -Dij.driver=3Dorg.apache.derby=
.jdbc.ClientDriver -Dij.connection.master=3Djdbc:derby://localhost:1527//ho=
me/os136789/Replication/testing/db_master/wombat -classpath /home/os136789/=
Apache/myDerbySandbox/trunk/jars/insane/derbyclient.jar:/home/os136789/Apac=
he/myDerbySandbox/trunk/jars/insane/derbytools.jar org.apache.derby.tools.i=
j /home/os136789/Replication/testing/verify_ReplicationTestRun.sql
> (Connection OK)
> Test run logfile.out plus derby.log from master and slave:
> Attachement: replicationLoad.tar=20
> From slave derby.log:
> ERROR XSLA7: Cannot redo operation null in the log.
> =09at org.apache.derby.iapi.error.StandardException.newException(Standard=
Exception.java:300)
> =09at org.apache.derby.impl.store.raw.log.FileLogger.redo(FileLogger.java=
:1525)
> =09at org.apache.derby.impl.store.raw.log.LogToFile.recover(LogToFile.jav=
a:920)
> =09at org.apache.derby.impl.store.raw.RawStore.boot(RawStore.java:334)
> =09at org.apache.derby.impl.services.monitor.BaseMonitor.boot(BaseMonitor=
.java:1999)
> =09at org.apache.derby.impl.services.monitor.TopService.bootModule(TopSer=
vice.java:291)
> =09at org.apache.derby.impl.services.monitor.BaseMonitor.startModule(Base=
Monitor.java:553)
> =09at org.apache.derby.iapi.services.monitor.Monitor.bootServiceModule(Mo=
nitor.java:419)
> =09at org.apache.derby.impl.store.access.RAMAccessManager.boot(RAMAccessM=
anager.java:1019)
> =09at org.apache.derby.impl.services.monitor.BaseMonitor.boot(BaseMonitor=
.java:1999)
> =09at org.apache.derby.impl.services.monitor.TopService.bootModule(TopSer=
vice.java:291)
> =09at org.apache.derby.impl.services.monitor.BaseMonitor.startModule(Base=
Monitor.java:553)
> =09at org.apache.derby.iapi.services.monitor.Monitor.bootServiceModule(Mo=
nitor.java:419)
> =09at org.apache.derby.impl.db.BasicDatabase.bootStore(BasicDatabase.java=
:793)
> =09at org.apache.derby.impl.db.BasicDatabase.boot(BasicDatabase.java:205)
> =09at org.apache.derby.impl.db.SlaveDatabase.bootBasicDatabase(SlaveDatab=
ase.java:425)
> =09at org.apache.derby.impl.db.SlaveDatabase.access$000(SlaveDatabase.jav=
a:70)
> =09at org.apache.derby.impl.db.SlaveDatabase$SlaveDatabaseBootThread.run(=
SlaveDatabase.java:312)
> =09at java.lang.Thread.run(Thread.java:619)
> Caused by: ERROR XSDFI: Error encountered while trying to write data to d=
isk during database recovery.  Check that the database disk is not full. If=
 it is then delete unnecessary files, and retry connecting to the database.=
  It is also possible that the file system is read only, or the disk has fa=
iled, or some other problem with the media.  System encountered error while=
 processing page Page(2,Container(0, 336)).
> B) No replication load.
> -------------------
> Master and slave servers started.
> Master: create master db.
> Master: freeze db
> Copy master db to slave.
> Master: unfreeze db.
> Slave:
> jdbc:derby://localhost:4527//home/os136789/Replication/testing/db_slave/w=
ombat;startSlave=3Dtrue;slaveHost=3Dlocalhost;slavePort=3D6666
> Master:
> jdbc:derby://localhost:1527//home/os136789/Replication/testing/db_master/=
wombat;startMaster=3Dtrue;slaveHost=3Dlocalhost;slavePort=3D6666
> Slave:
> SQLException: -1 XRE08 DERBY SQL error: SQLCODE: -1, SQLSTATE: XRE08, SQL=
ERRMC: Replication slave mode started successfully for database '/home/os13=
6789/Replication/testing/db_slave/wombat'. Connection refused because the d=
atabase is in replication slave mode.
> NOT Running tests against master.
> Master: jdbc:derby://localhost:1527//home/os136789/Replication/testing/db=
_master/wombat;failover=3Dtrue
> failOver_direct Got SQLException: -1 XRE20 DERBY SQL error: SQLCODE: -1, =
SQLSTATE: XRE20, SQLERRMC: Failover performed successfully for database '/h=
ome/os136789/Replication/testing/db_master/wombat', the database has been s=
hutdown.
> Slave:
> /usr/local/java/jdk1.6/jre/lib/../bin/java -Dij.driver=3Dorg.apache.derby=
.jdbc.ClientDriver -Dij.connection.slave=3Djdbc:derby://localhost:4527//hom=
e/os136789/Replication/testing/db_slave/wombat -classpath /home/os136789/Ap=
ache/myDerbySandbox/trunk/jars/insane/derbyclient.jar:/home/os136789/Apache=
/myDerbySandbox/trunk/jars/insane/derbytools.jar org.apache.derby.tools.ij =
/home/os136789/Replication/testing/verify_ReplicationTestRun.sql
> (Connection OK)
> Master:
> /usr/local/java/jdk1.6/jre/lib/../bin/java -Dij.driver=3Dorg.apache.derby=
.jdbc.ClientDriver -Dij.connection.master=3Djdbc:derby://localhost:1527//ho=
me/os136789/Replication/testing/db_master/wombat -classpath /home/os136789/=
Apache/myDerbySandbox/trunk/jars/insane/derbyclient.jar:/home/os136789/Apac=
he/myDerbySandbox/trunk/jars/insane/derbytools.jar org.apache.derby.tools.i=
j /home/os136789/Replication/testing/verify_ReplicationTestRun.sql
> Failure: Hangs! (On 'show schemas;')=20
> Test run logfile.out plus derby.log from master and slave:
> Attachement: NOreplicationLoad.tar=20

--=20
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.