[ https://issues.apache.org/jira/browse/AMQ-6522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15705032#comment-15705032 ]
Gary Tully commented on AMQ-6522: --------------------------------- symptom with a message > 32k in size{code} 14:23:16,374 | ERROR | AMQ-1-thread-1 | ActiveMQServiceFactory | Exception on start: java.io.IOException: Detected missing/corrupt journal files referenced by:[xxx] 1 messages affected. java.io.IOException: Detected missing/corrupt journal files referenced by:[0:TEST] 1 messages affected. {code} > KahaDB - checkForCorruptJournalFiles has hard coded 32k batch size limit in > error > ---------------------------------------------------------------------------------- > > Key: AMQ-6522 > URL: https://issues.apache.org/jira/browse/AMQ-6522 > Project: ActiveMQ > Issue Type: Bug > Components: KahaDB, Message Store > Affects Versions: 5.14.0 > Reporter: Gary Tully > Assignee: Gary Tully > Fix For: 5.15.0 > > > Using setCheckForCorruptJournalFiles=true, there is a sanity check of the > journal at startup - skipping through each of the journal files validating > batch record checksums and magic in an effort to detect corruption early and > drop the relevant messages or error out. > There is an error in the check logic that assumes a batch record must be < > 32k which is not the case if a message is > 32k > Tidy up this check and add some more detail to the error messages. -- This message was sent by Atlassian JIRA (v6.3.4#6332)