[jira] [Updated] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jan Eerdekens updated LUCENE-6725: -- Attachment: branch5-jdk8u51-results.txt branch5-jdk7u75-results.txt I've been only able to run 2 tests, using different JVMs, for the 5 branch. I first tried to run all tests with all the params you gave, but those took very long 30+ minutes and eventually the machine ran out of disk space. The DEV machine is a bit memory and disk constraint, so I re-ran the tests, but this time only test-core without the params which took about 10 minutes to run each time. Just let me know if this is sufficient. If so I should also be able to run to same for trunk on Thursday. > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: branch5-jdk7u75-results.txt, > branch5-jdk8u51-results.txt, hs_err_pid18938.log, > lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14679731#comment-14679731 ] Jan Eerdekens commented on LUCENE-6725: --- [~dawid.weiss]: I can spare a couple of hours today on the machine when I don't need it for testing. What's the difference in running time between the full set and the core tests alone? > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log, lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14679713#comment-14679713 ] Jan Eerdekens commented on LUCENE-6725: --- [~dawid.weiss]: today and maybe Thursday too I have access to the Solaris machine so I can run the testsuites for 5 and trunk if that can help you for other things than my issue (as that seems to be purely a JVM + Lucene 3.5 related issue, which we can work around by upgrading the JVM)? If they can help you, what results/files should I gather and send to you? > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log, lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14661820#comment-14661820 ] Jan Eerdekens commented on LUCENE-6725: --- The tests for Lucene 4.10.4 for the 3 JVM versions don't seem to cause JVM crashes. > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log, lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14661573#comment-14661573 ] Jan Eerdekens commented on LUCENE-6725: --- A next batch of results, still for 3.5 (newer is next): on 7u80 I still get the same crashes, but on 8u51 all the tests succeed and I see no crashes. It is really starting to look like a JVM issue (and probably really specific to Solaris/SPARC). A JVM upgrade is an option, albeit that JDK8 is only 'supported' by Liferay EE, but not 'certified' yet. I'll try to do some tests with Lucene 4.10 and maybe also already try to start Liferay with the JDK8 version and see how it behaves. > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log, lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Updated] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jan Eerdekens updated LUCENE-6725: -- Attachment: lucene-3.5-ant-test-results.txt My first test run, for Lucene 3.5 on our test environment, already causes the same JVM crashes (SIGSEGV on IndexFileNames.segmentFileName) as we saw when triggering a reindex in Liferay on production. What files should I provide: just the console output of the test, all the hs_err_pid files or even the whole test subdirectory? > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log, lucene-3.5-ant-test-results.txt > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659954#comment-14659954 ] Jan Eerdekens commented on LUCENE-6725: --- I have only direct machine access to the test environment and will run the tests the next time I'm at the client. I've asked them to do at least the Lucene 3.5 and Lucene 4 tests on the pre-production and production environments and send me the results. I don't know what the timing will be for that as they'll need to schedule someone to do this during what is currently a vacation period over here. Also the fact that the tests try to download stuff from the internet will possibly complicate the installation and running a bit. Thx for the help already and hopefully I'll be able to get back to you soon with some results. > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659867#comment-14659867 ] Jan Eerdekens commented on LUCENE-6725: --- Yup, but sadly we're stuck with that version as it is bundled with our Liferay version (which is a recent one... but uses an ancient Lucene version). I think we'll have to switch to a separate SOLR server to also solve some performance issues and that would allow us to use newer SOLR/Lucene versions. To be sure I'll have them run the tests for both versions so we gather the most data. > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659855#comment-14659855 ] Jan Eerdekens commented on LUCENE-6725: --- [~dawid.weiss]: we also contacted Liferay support and their conclusion was also that it'll probably be some sort of JVM issue, but I just wanted to see if you guys thought the same. Running the Lucene test suite is a great idea that I didn't think of. I'll try to get someone of the admin team to run it on the production server and see what result (and hopefully crashes) that gives. Is it enough to download the Lucene 3.5.0 sources and run 'ant test' or are you referring to other tests? > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659794#comment-14659794 ] Jan Eerdekens commented on LUCENE-6725: --- Problem seems (very) similar to what this guy reported in the mailing list: http://mail-archives.apache.org/mod_mbox/lucene-java-user/201306.mbox/%3CCA+9jHe=H9-wun5FedByRr5GuKKt1O7f=xh3nvgbikkb8wt1...@mail.gmail.com%3E > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Commented] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14659774#comment-14659774 ] Jan Eerdekens commented on LUCENE-6725: --- Currently reproducing this behaviour is difficult as we're seeing the following: - the problem seemed to occur consistently on our production environment (but we can't tried it a lot there for obvious reasons) - after installing a monitoring tool today (a Java agent) the next reindex didn't crash the JVM - the problem (at first) didn't happen on our pre-production environment (which is the most similar with regards to CPU/memory/... to production) - in the weeks after discovering the problem, it also happened once on the pre-production environment, but on retrying the reindex afterwards it didn't happen again - the problem isn't reproducible on our test environment or local machines (Apple MacBook Pro's) We also created a small standalone Java tool that is also able to do the IndexWriter.deleteAll() call on a copy of the index, but that doesn't trigger the same JVM crash on our production and pre-production environments (using in the same JVM version). > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Minor > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Updated] (LUCENE-6725) Reindex crashes the JVM
[ https://issues.apache.org/jira/browse/LUCENE-6725?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jan Eerdekens updated LUCENE-6725: -- Attachment: hs_err_pid18938.log > Reindex crashes the JVM > --- > > Key: LUCENE-6725 > URL: https://issues.apache.org/jira/browse/LUCENE-6725 > Project: Lucene - Core > Issue Type: Bug > Components: core/index >Affects Versions: 3.5 > Environment: Solaris 10 1/13 (Update 11) Patchset applied. > Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC > CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus > Memory: 8k page, physical 25165824k(3240848k free) > vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE > (1.7.0_75-b13) >Reporter: Jan Eerdekens >Priority: Blocker > Attachments: hs_err_pid18938.log > > > We're using Liferay which uses Lucene behind the screens to index things like > documents, web content, users, etc... . When we trigger a full reindex via > the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM > crashes and generates a dump with the following message: > # > # A fatal error has been detected by the Java Runtime Environment: > # > # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 > # > # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build > 1.7.0_75-b13) > # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode > solaris-sparc compressed oops) > # Problematic frame: > # J 5227 C2 > org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; > (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
[jira] [Created] (LUCENE-6725) Reindex crashes the JVM
Jan Eerdekens created LUCENE-6725: - Summary: Reindex crashes the JVM Key: LUCENE-6725 URL: https://issues.apache.org/jira/browse/LUCENE-6725 Project: Lucene - Core Issue Type: Bug Components: core/index Affects Versions: 3.5 Environment: Solaris 10 1/13 (Update 11) Patchset applied. Oracle Solaris 10 8/11 s10s_u10wos_17b SPARC CPU:total 64 v9, popc, vis1, vis2, vis3, blk_init, cbcond, sun4v, niagara_plus Memory: 8k page, physical 25165824k(3240848k free) vm_info: Java HotSpot(TM) 64-Bit Server VM (24.75-b04) for solaris-sparc JRE (1.7.0_75-b13) Reporter: Jan Eerdekens Priority: Blocker We're using Liferay which uses Lucene behind the screens to index things like documents, web content, users, etc... . When we trigger a full reindex via the Liferay Control Panel, which uses IndexWriter.deleteAll(), the JVM crashes and generates a dump with the following message: # # A fatal error has been detected by the Java Runtime Environment: # # SIGSEGV (0xb) at pc=0x78de94a8, pid=18938, tid=2478 # # JRE version: Java(TM) SE Runtime Environment (7.0_75-b13) (build 1.7.0_75-b13) # Java VM: Java HotSpot(TM) 64-Bit Server VM (24.75-b04 mixed mode solaris-sparc compressed oops) # Problematic frame: # J 5227 C2 org.apache.lucene.index.IndexFileNames.segmentFileName(Ljava/lang/String;Ljava/lang/String;)Ljava/lang/String; (44 bytes) @ 0x78de94a8 [0x78de9480+0x28] -- This message was sent by Atlassian JIRA (v6.3.4#6332) - To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org
