[jira] [Commented] (HBASE-24) Scaling: Too many open file handles to datanodes
[ https://issues.apache.org/jira/browse/HBASE-24?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14267759#comment-14267759 ] Cosmin Lehene commented on HBASE-24: [~saint@gmail.com] this is a venerable one from the oldies but goldies series :) as we long ago learnt to up xceivers max open file handles and the like, we haven't been affected, however I'm curious if this is something that would be worth a refresh. > Scaling: Too many open file handles to datanodes > > > Key: HBASE-24 > URL: https://issues.apache.org/jira/browse/HBASE-24 > Project: HBase > Issue Type: Bug > Components: regionserver >Reporter: stack >Priority: Blocker > Attachments: HBASE-823.patch, MonitoredReader.java > > > We've been here before (HADOOP-2341). > Today the rapleaf gave me an lsof listing from a regionserver. Had thousands > of open sockets to datanodes all in ESTABLISHED and CLOSE_WAIT state. On > average they seem to have about ten file descriptors/sockets open per region > (They have 3 column families IIRC. Per family, can have between 1-5 or so > mapfiles open per family -- 3 is max... but compacting we open a new one, > etc.). > They have thousands of regions. 400 regions -- ~100G, which is not that > much -- takes about 4k open file handles. > If they want a regionserver to server a decent disk worths -- 300-400G -- > then thats maybe 1600 regions... 16k file handles. If more than just 3 > column families. then we are in danger of blowing out limits if they are > 32k. > We've been here before with HADOOP-2341. > A dfsclient that used non-blocking i/o would help applications like hbase > (The datanode doesn't have this problem as bad -- CLOSE_WAIT on regionserver > side, the bulk of the open fds in the rapleaf log, don't have a corresponding > open resource on datanode end). > Could also just open mapfiles as needed, but that'd kill our random read > performance and its bad enough already. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-24) Scaling: Too many open file handles to datanodes
[ https://issues.apache.org/jira/browse/HBASE-24?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13039507#comment-13039507 ] Ted Yu commented on HBASE-24: - Should we take further action on this JIRA ? > Scaling: Too many open file handles to datanodes > > > Key: HBASE-24 > URL: https://issues.apache.org/jira/browse/HBASE-24 > Project: HBase > Issue Type: Bug > Components: regionserver >Reporter: stack >Priority: Blocker > Fix For: 0.92.0 > > Attachments: HBASE-823.patch, MonitoredReader.java > > > We've been here before (HADOOP-2341). > Today the rapleaf gave me an lsof listing from a regionserver. Had thousands > of open sockets to datanodes all in ESTABLISHED and CLOSE_WAIT state. On > average they seem to have about ten file descriptors/sockets open per region > (They have 3 column families IIRC. Per family, can have between 1-5 or so > mapfiles open per family -- 3 is max... but compacting we open a new one, > etc.). > They have thousands of regions. 400 regions -- ~100G, which is not that > much -- takes about 4k open file handles. > If they want a regionserver to server a decent disk worths -- 300-400G -- > then thats maybe 1600 regions... 16k file handles. If more than just 3 > column families. then we are in danger of blowing out limits if they are > 32k. > We've been here before with HADOOP-2341. > A dfsclient that used non-blocking i/o would help applications like hbase > (The datanode doesn't have this problem as bad -- CLOSE_WAIT on regionserver > side, the bulk of the open fds in the rapleaf log, don't have a corresponding > open resource on datanode end). > Could also just open mapfiles as needed, but that'd kill our random read > performance and its bad enough already. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-24) Scaling: Too many open file handles to datanodes
[ https://issues.apache.org/jira/browse/HBASE-24?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13089042#comment-13089042 ] Alex Newman commented on HBASE-24: -- +conf); // defer opening streams is in the patch. Are we actually still defering. That seems wrong. > Scaling: Too many open file handles to datanodes > > > Key: HBASE-24 > URL: https://issues.apache.org/jira/browse/HBASE-24 > Project: HBase > Issue Type: Bug > Components: regionserver >Reporter: stack >Priority: Blocker > Attachments: HBASE-823.patch, MonitoredReader.java > > > We've been here before (HADOOP-2341). > Today the rapleaf gave me an lsof listing from a regionserver. Had thousands > of open sockets to datanodes all in ESTABLISHED and CLOSE_WAIT state. On > average they seem to have about ten file descriptors/sockets open per region > (They have 3 column families IIRC. Per family, can have between 1-5 or so > mapfiles open per family -- 3 is max... but compacting we open a new one, > etc.). > They have thousands of regions. 400 regions -- ~100G, which is not that > much -- takes about 4k open file handles. > If they want a regionserver to server a decent disk worths -- 300-400G -- > then thats maybe 1600 regions... 16k file handles. If more than just 3 > column families. then we are in danger of blowing out limits if they are > 32k. > We've been here before with HADOOP-2341. > A dfsclient that used non-blocking i/o would help applications like hbase > (The datanode doesn't have this problem as bad -- CLOSE_WAIT on regionserver > side, the bulk of the open fds in the rapleaf log, don't have a corresponding > open resource on datanode end). > Could also just open mapfiles as needed, but that'd kill our random read > performance and its bad enough already. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira