[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835578#comment-13835578 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in HBase-0.94-security #349 (See [https://builds.apache.org/job/HBase-0.94-security/349/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546659) * /hbase/branches/0.94/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835585#comment-13835585 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in HBase-0.94 #1215 (See [https://builds.apache.org/job/HBase-0.94/1215/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546659) * /hbase/branches/0.94/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835588#comment-13835588 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in HBase-TRUNK #4703 (See [https://builds.apache.org/job/HBase-TRUNK/4703/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546658) * /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835591#comment-13835591 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in hbase-0.96-hadoop2 #137 (See [https://builds.apache.org/job/hbase-0.96-hadoop2/137/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546660) * /hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835594#comment-13835594 ] Hudson commented on HBASE-9915: --- FAILURE: Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #857 (See [https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/857/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546658) * /hbase/trunk/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13835602#comment-13835602 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in hbase-0.96 #208 (See [https://builds.apache.org/job/hbase-0.96/208/]) HBASE-10058 Test for HBASE-9915 (avoid reading index blocks) (larsh: rev 1546660) * /hbase/branches/0.96/hbase-server/src/test/java/org/apache/hadoop/hbase/regionserver/TestBlocksScanned.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817216#comment-13817216 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in hbase-0.96 #184 (See [https://builds.apache.org/job/hbase-0.96/184/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539934) * /hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817230#comment-13817230 ] Hudson commented on HBASE-9915: --- FAILURE: Integrated in HBase-TRUNK #4674 (See [https://builds.apache.org/job/HBase-TRUNK/4674/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539933) * /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817235#comment-13817235 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in HBase-0.94-security #331 (See [https://builds.apache.org/job/HBase-0.94-security/331/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539936) * /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817255#comment-13817255 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in HBase-0.94 #1197 (See [https://builds.apache.org/job/HBase-0.94/1197/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539936) * /hbase/branches/0.94/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817381#comment-13817381 ] Hudson commented on HBASE-9915: --- SUCCESS: Integrated in hbase-0.96-hadoop2 #116 (See [https://builds.apache.org/job/hbase-0.96-hadoop2/116/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539934) * /hbase/branches/0.96/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817484#comment-13817484 ] Hudson commented on HBASE-9915: --- FAILURE: Integrated in HBase-TRUNK-on-Hadoop-2.0.0 #832 (See [https://builds.apache.org/job/HBase-TRUNK-on-Hadoop-2.0.0/832/]) HBASE-9915 Performance: isSeeked() in EncodedScannerV2 always returns false (larsh: rev 1539933) * /hbase/trunk/hbase-server/src/main/java/org/apache/hadoop/hbase/io/hfile/HFileReaderV2.java Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817090#comment-13817090 ] James Taylor commented on HBASE-9915: - Did I hear someone call that it's time to wrap-up for the 0.94.14 release :-). Awesome work, Lars! Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)
[jira] [Commented] (HBASE-9915) Performance: isSeeked() in EncodedScannerV2 always returns false
[ https://issues.apache.org/jira/browse/HBASE-9915?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13817092#comment-13817092 ] Lars Hofhansl commented on HBASE-9915: -- Since 0.94.13 was somewhat late, I'm planning to do 0.94.14 a bit sooner; before the end of the month. :) Performance: isSeeked() in EncodedScannerV2 always returns false Key: HBASE-9915 URL: https://issues.apache.org/jira/browse/HBASE-9915 Project: HBase Issue Type: Bug Components: Scanners Reporter: Lars Hofhansl Assignee: Lars Hofhansl Labels: performance Fix For: 0.98.0, 0.96.1, 0.94.14 Attachments: 9915-0.94.txt, 9915-trunk-v2.txt, 9915-trunk-v2.txt, 9915-trunk.txt, profile.png While debugging why reseek is so slow I found that it is quite broken for encoded scanners. The problem is this: AbstractScannerV2.reseekTo(...) calls isSeeked() to check whether scanner was seeked or not. If it was it checks whether the KV we want to seek to is in the current block, if not it always consults the index blocks again. isSeeked checks the blockBuffer member, which is not used by EncodedScannerV2 and thus always returns false, which in turns causes an index lookup for each reseek. -- This message was sent by Atlassian JIRA (v6.1#6144)