Clone Tools
Constraints: committers
Constraints: files
Constraints: dates
dop check-in rework.

Change-Id: Idb21dfdbdff76228f9634abc0d1feac6e9e1810b

    • -11
    • +22
Merge "Improved control point logging"

Typos and misspellings fix 4

Corrected spelling of "disappeared" in 26 user messages in odbc

Change-Id: Ib7613a49aa05a93125d3769d4cdbf3d58ea4356c

Merge "SQL Compiler/Generator speed improvement"

Catalog api functionality for T2 driver

Change-Id: Iee17029332f2815279b96f6cedf8fc75d042d7fb

    • -4
    • +0
    • -34
    • +6
    • -23
    • +3
    • -38
    • +7
    • -26
    • +541
    • -7
    • +38
    • -9
    • +1
    • -5
    • +3
    • -1791
    • +65
    • -240
    • +237
Merge "Fix for _DTM_ schema, Launchpad 1365202"

Merge "Fix for _DTM_ schema, Launchpad 1365202"

Merge "Changing the default JVM heap size to 1GB"

Fix for _DTM_ schema, Launchpad 1365202

Change-Id: Id554b2fb87f75a97800cc47a2082b81ab7d07a49

    • -7
    • +18
    • -14
    • +11
Merge "Allow RMS shared segment without huge pages"

Squashed commit of the following:

commit 221d4199001b3f06d5629b82ed1281bdfb95f043

Author: qchen <>

Date: Mon Sep 8 19:32:05 2014 +0000

DoP: use a better version of resource estimator

Change-Id: Idf6ea5caa7e4915c65b4d54f58d7e483c26871f4

commit 75c92f255ad3841102757f6188d749db42eaedf3

Author: qchen <>

Date: Fri Sep 5 15:55:11 2014 +0000

do not add ESP partition requirement to partial GB leaf1 and leaf2.

commit f7c652adf2953d18f10da4a952f78d3815b5b3d4

Author: qchen <>

Date: Fri Sep 5 15:20:40 2014 +0000


commit a419954b1528ca0d1f84c5298b806b46f561a713

Author: qchen <>

Date: Wed Sep 3 23:45:13 2014 +0000

add a check on stats smart ptr before its use.

commit 8a2dcc24f0081d554ec3d69560c050f720da04dd

Author: qchen <>

Date: Wed Sep 3 18:31:50 2014 +0000

partial gb root running in Master; use rowcount from stats for Hbase

in AppliedStatMan::getStatsForCANodeId().

commit 55f46ededa9fcb30d38d17be6692e1535b4a274c

Author: qchen <>

Date: Fri Aug 29 17:46:00 2014 +0000

improve the logic to disable parallel GB partial root when rows < 5000

commit 743783720d030f38e50a065737c8c218fa47c6b5

Author: Ravisha Neelakanthappa <>

Date: Wed Aug 20 17:20:09 2014 +0000

Fix for bug 1348317. Enable Adaptive Segmentation logic to

compute DoP based on resource estimation.

Change-Id: I7c2b0756854d5a86d8bde162dad064b4e924233c

Change-Id: Id68e1d8265e478f128ed61d1cabc4849873dae23

    • -37
    • +72
  1. … 5 more files in changeset.
Changing the default JVM heap size to 1GB

The default heap size for JVM in the Trafodion processes is changed to 1GB

from 512MB. Earlier attempt to change it was incomplete.

Change-Id: I7de2ee1ee8cd60a2c0ebb3067baba1181dd4f210

Merge "Bulk Extract fixes"

Merge "Fix bug #1365658 - Eliminate absolute pathname to"

Fix for _DTM_ schema, Launchpad 1365202

initialize trafodion, drop will no longer drop dtm log files.

Also, one can no longer create a user table in schema "_DTM_". This

is a reserved schema name for dtm log files.

Change-Id: I50b66303d9aea20e35b9d8fc4c085aff449163d9

    • -0
    • +11
    • -5
    • +10
Add ability to check for absolute files used as dlls

Added script and makefile target for detecting absolute filenames as

seen in bug 1365658. The make target is not yet used in Jenkins

builds. Because the check takes more than 30 seconds, it was not

added as part of the normal "make package" work, a 20+% increase in

running make is not worthwhile for a rare bug which may never happen



$ time make find-absolute-dlls


Searching <my-directory>

for absolute files used as dlls ...

ERROR: Found 6 files with absolute filenames as dlls:


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/


NEEDED /home/tools/jdk1.7.0_09_64/jre/lib/amd64/server/

make: *** [find-absolute-dlls] Error 6

real 0m39.801s

user 0m26.196s

sys 0m9.438s

Also made a small cleanup to the existing check_man script.

Second patch set added -v to help text.

Change-Id: I22babe08dbf8a4620603fc670acc6b05e831222f

    • -0
    • +102
Allow RMS shared segment without huge pages

This change allows creation and use of the RMS shared

memory segment in cases where huge pages are not available.

If huge pages are available, RMS will use them as before

this change.

Change-Id: I81909a5ac968a57d66bdcdaf172af470c585bf5c

Closes-Bug: #1366033

Bulk Extract fixes

Change-Id: Ic2d6369c0dfc61cd647715befa810c0442124d91

Merge "Pre-fetch cells from Hbase"

Merge "Fix bug #1328254 - Explain options 'm' got TRANSLATE error"

Merge "Fix SIGSEGV in ExLobsOper"

Merge "Related to bug 1335477: Improve debugging code for MDAM."

Merge "Update release version to 0.9.0"

Fix bug #1365658 - Eliminate absolute pathname to

Changes the makefiles used to link sqlci, tdm_arkcmp, etc. so that is referenced generically rather than by absolute

pathname. This will allow us (at run time) to use

from whereever it has been installed ... much like the way

that we get to libjvm.

All dev regressions were run to ensure that the fix has no


Also, verified (using objdump) that there is no longer any

reference to the absolute pathname for in the

built objects for sqlci, tdm_arkcmp, tdm_arkesp, tdm_udrserv,

mxssmp, or mxsscp.

Files changed:



Change-Id: I40d1cf2e295871e663a55e198b1465bcb81b5655

Pre-fetch cells from Hbase

Pre-fetch is enabled via a parameter in HTableClient.startScan method.Pre-fetch

is not done for unique and batch Trafodion operations and all native

Hbase table access. Pre-fetch is currently disabled for non-unique UMD

Trafodion operations.

startScan method invokes pre-fetch to Hbase in a different thread. When the

fetchRows method is called, pre-fetch completes, passes cell info to JNI and

invokes pre-fetch if there are more rows to be fetched.

We have observed around 45% reduction in response time to fetch 12 million

rows of a sizteen partition table in a node via a single process.

Change-Id: I3c81e182663fddd08a2fc873a39302b179850c92

    • -11
    • +23
    • -84
    • +125
Bulk Unload (prototype) fixes

fixes for the parser and other issues

Change-Id: If0e3a508bf45c9b34d84083e4fb3906734b5db73

    • -8
    • +11
ODBC win64 driver compression ability

Change-Id: I2ba5c7e82f8162955720dc6360bb5ea3371d7ed3

    • -0
    • +118
    • -0
    • +46
    • -14
    • +10
    • -18
    • +1
  1. … 16 more files in changeset.
Related to bug 1335477: Improve debugging code for MDAM.

If the Trafodion code is built using NA_MDAM_EXECUTOR_DEBUG, then

whenever a query uses an MDAM plan, the MDAM network will be

printed to cout when the query executes. This can be useful for

debugging MDAM queries.

Formerly, the code printed out pointers to the location of the

data values (and with the 64-bit conversion the pointers are

badly mangled). This is nearly useless for debugging.

With these changes, the code assumes the key column is a CHAR(10)

and prints out the actual value. (Note that there is no easy way

at present of determining the actual data type. If the data isn't

a CHAR(10), you'll get what you get; garbage most likely. But

since you're debugging anyway you can change the code to take

into account the actual data type of the example you are


Change-Id: I2d53159c1b10e28777916850efc3f325d3ddb234

Improved control point logging

Change-Id: I0568ad437d3c2f89b49ba8c5c9503e80014c4c3f

Merge "Fix for #1359346 Java version detection"