[
https://issues.apache.org/jira/browse/IGNITE-17665?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17605298#comment-17605298
]
Yury Gerzhedovich commented on IGNITE-17665:
--------------------------------------------
[~alexey.makarevich] it's not real B+Tree corruption, unhandled exceptions
converted to CorruptedTreeException. So, after fixing client code, node could
be up and running.
Of course, that doesn't change the fact that this is a bug and should be fixed
> Ignite node crash(CorruptedTreeException: B+Tree is corrupted) when using
> JDBC Thin driver
> ------------------------------------------------------------------------------------------
>
> Key: IGNITE-17665
> URL: https://issues.apache.org/jira/browse/IGNITE-17665
> Project: Ignite
> Issue Type: Bug
> Components: jdbc, sql
> Affects Versions: 2.12, 2.13
> Reporter: Alexey Makarevich
> Priority: Critical
> Attachments:
> ignite-4986b884.0-partitioned-distributedJoins-CRASH-String.log,
> ignite-e4793913.0-replicated-CRASH-String.log
>
>
> Create test table
> {code:java}
> CREATE TABLE `Table1` (
> `tid` INT PRIMARY KEY,
> `descr` VARCHAR NOT NULL
> -- ) WITH "template=replicated, CACHE_NAME=Table1";
> ) WITH "CACHE_NAME=Table1";
> INSERT INTO Table1 (tid,descr) VALUES (236,'CCCcCcccCc'); {code}
> Try to select it with JDBC thin driver (ignite-core-2.13.0.jar for 2.13)
> {code:java}
> Class.forName("org.apache.ignite.IgniteJdbcThinDriver");
> Connection conn =
> DriverManager.getConnection("jdbc:ignite:thin://192.168.56.101:10800");
> PreparedStatement stmt = conn.prepareStatement("select * from TABLE1 WHERE
> TID=?");
> stmt.setString(1, "asdhjhds");
> ResultSet rs = stmt.executeQuery();{code}
> In this case we receive
> {code:java}
> java.sql.SQLException: class org.apache.ignite.IgniteCheckedException: Data
> conversion error converting "asdhjhds" [22018-197] {code}
> But if we get connection with enabled distributedJoins
> {code:java}
> Connection conn =
> DriverManager.getConnection("jdbc:ignite:thin://192.168.56.101:10800;distributedJoins=true");
> {code}
> there will be crash
> {code:java}
> [20:00:34,717][FINE][client-connector-#91][ClientListenerNioListener] Client
> request received [reqId=1, addr=/192.168.56.1:12755,
> req=JdbcQueryExecuteRequest [schemaName=PUBLIC, pageSize=1024, maxRows=0,
> sqlQry=select * from TABLE1 WHERE TID=?, args=Object[] [asdhjhds],
> stmtType=SELECT_STATEMENT_TYPE, autoCommit=true, partResReq=false,
> explicitTimeout=false, super=JdbcRequest [type=2, reqId=1]]]
> [20:00:34,763][SEVERE][client-connector-#91][] Critical system error
> detected. Will be handled accordingly to configured handler
> [hnd=StopNodeOrHaltFailureHandler [tryStop=false, timeout=0,
> super=AbstractFailureHandler [ignoredFailureTypes=UnmodifiableSet
> [SYSTEM_WORKER_BLOCKED, SYSTEM_CRITICAL_OPERATION_TIMEOUT]]],
> failureCtx=FailureContext [type=CRITICAL_ERROR, err=class
> o.a.i.i.processors.cache.persistence.tree.CorruptedTreeException: B+Tree is
> corrupted [groupId=-1797510589, pageIds=[844420635164673],
> cacheId=-1797510589, cacheName=Table1, indexName=_key_PK, msg=Runtime failure
> on lookup row: IndexSearchRowImpl
> [rowHnd=o.a.i.i.processors.query.h2.index.QueryIndexRowHandler@2c34fb40]]]]class
>
> org.apache.ignite.internal.processors.cache.persistence.tree.CorruptedTreeException:
> B+Tree is corrupted [groupId=-1797510589, pageIds=[844420635164673],
> cacheId=-1797510589, cacheName=Table1, indexName=_key_PK, msg=Runtime failure
> on lookup row: IndexSearchRowImpl
> [rowHnd=org.apache.ignite.internal.processors.query.h2.index.QueryIndexRowHandler@2c34fb40]]
> {code}
> The same crash will be if we create table in replicated mode "WITH
> "template=replicated,".
> It seems that in crash cases there are no type validations\conversions...
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
