[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909118#comment-14909118 ] Lars Hofhansl commented on HBASE-14489: --- Great. Thanks. I'll commit some time tomorrow. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909540#comment-14909540 ] Hudson commented on HBASE-14489: SUCCESS: Integrated in HBase-1.2-IT #172 (See [https://builds.apache.org/job/HBase-1.2-IT/172/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 3633380e2491d503b59f72c4bbc6ef5db8b261e6) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909527#comment-14909527 ] Lars Hofhansl commented on HBASE-14489: --- Committed to a million branches: 2.x, 1.x, 1.2, 1.1, 1.0, 0.98. Thanks for the review. [~giacomotaylor], [~mujtabachohan], maybe we can do a pherf run on this? Need current 0.98 branch HBase. It would show up in queries that filters most rows at the server. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909545#comment-14909545 ] Hudson commented on HBASE-14489: SUCCESS: Integrated in HBase-1.3-IT #187 (See [https://builds.apache.org/job/HBase-1.3-IT/187/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 45fdcfc23ca6e9bf8f5e50eb15bb9beee7bd6c27) * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909555#comment-14909555 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-0.98 #1132 (See [https://builds.apache.org/job/HBase-0.98/1132/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev e199e1e984dfec0c91f583e3ce3968c30983acde) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909560#comment-14909560 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-1.2 #205 (See [https://builds.apache.org/job/HBase-1.2/205/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 3633380e2491d503b59f72c4bbc6ef5db8b261e6) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909574#comment-14909574 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-TRUNK #6846 (See [https://builds.apache.org/job/HBase-TRUNK/6846/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 526520de0a9d7a29fcf1b4c521f017ca75a46cbc) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909556#comment-14909556 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-1.0 #1060 (See [https://builds.apache.org/job/HBase-1.0/1060/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 8ea34aea023952df21f6ede66b63536960400ec1) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909584#comment-14909584 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-0.98-on-Hadoop-1.1 #1085 (See [https://builds.apache.org/job/HBase-0.98-on-Hadoop-1.1/1085/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev e199e1e984dfec0c91f583e3ce3968c30983acde) * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909568#comment-14909568 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-1.1 #683 (See [https://builds.apache.org/job/HBase-1.1/683/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 8dba3e8905454d7547dfd04d880687c544fefa0a) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909573#comment-14909573 ] Hudson commented on HBASE-14489: FAILURE: Integrated in HBase-1.3 #210 (See [https://builds.apache.org/job/HBase-1.3/210/]) HBASE-14489 postScannerFilterRow consumes a lot of CPU. (larsh: rev 45fdcfc23ca6e9bf8f5e50eb15bb9beee7bd6c27) * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/SimpleRegionObserver.java * hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/RegionCoprocessorHost.java * hbase-server/src/test/java/org/apache/hadoop/hbase/coprocessor/TestRegionObserverInterface.java > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt, 14489-master.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909086#comment-14909086 ] Lars Hofhansl commented on HBASE-14489: --- Thanks [~apurtell]. Is this OK: {code} if (clazz == BaseRegionObserver.class) { {code} ? Can I compare classes by reference like this with all the classloader huh hah we're doing? > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909088#comment-14909088 ] Lars Hofhansl commented on HBASE-14489: --- By the way, testing this with Phoenix, which does add a coprocessor to each table... A query that filters most rows out I also see a ~2.5x performance improvement. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14909114#comment-14909114 ] Andrew Purtell commented on HBASE-14489: Yes that equality test is safe because the coprocessor classloader delegates to the parent for org.apache.hbase.* and select others, which will be the same one used to load the regionserver's classes. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl >Assignee: Lars Hofhansl > Labels: performance > Fix For: 2.0.0, 1.2.0, 1.3.0, 0.98.15, 1.0.3, 1.1.3 > > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14908948#comment-14908948 ] Andrew Purtell commented on HBASE-14489: +1 > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl > Labels: performance > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14907666#comment-14907666 ] Andrew Purtell commented on HBASE-14489: I'll look at the patch when at the desktop tomorrow. I do remember microbenchmarking the classloader checks when we added them. You can find the results on the jira that confirm your suspicions. It was a necessary change but then hook calls were no longer 'free'. Even a coprocessor that does nothing per hook will accrue that overhead and early out won't work because something is installed. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl > Labels: performance > Attachments: 14489-0.98.txt > > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HBASE-14489) postScannerFilterRow consumes a lot of CPU
[ https://issues.apache.org/jira/browse/HBASE-14489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14907576#comment-14907576 ] Lars Hofhansl commented on HBASE-14489: --- Hmm... Also see HBASE-10047 where I filed the same thing, even with a patch. I am reviving the patch here. > postScannerFilterRow consumes a lot of CPU > -- > > Key: HBASE-14489 > URL: https://issues.apache.org/jira/browse/HBASE-14489 > Project: HBase > Issue Type: Bug >Reporter: Lars Hofhansl > Labels: performance > > During an unrelated test I found that when scanning a tall table with CQ only > and filtering most results at the server, 50%(!) of time is spend in > postScannerFilterRow, even though the coprocessor does nothing in that hook. > We need to find a way not to call this hook when not needed, or to question > why we have this hook at all. > I think [~ram_krish] added the hook (or maybe [~anoop.hbase]). I am also not > sure whether Phoenix uses this hook ([~giacomotaylor]?) -- This message was sent by Atlassian JIRA (v6.3.4#6332)