1. Input and output RowGroup's used in GROUP_CONCAT classes
are currently allocating a raw memory buffer of size equal
to the actual width of the string datatype. As an example,
for the following query:
SELECT col1, GROUP_CONCAT(col2) FROM t GROUP BY col1;
If col2 is a TEXT field with default width, the input
RowGroup containing the target rows to be concatenated will
assign 64kb of memory for every input row in the RowGroup.
This is wasteful as actual field values in real workloads
would be much smaller. We fix this by enabling the
RowGroup to use the StringStore when the RowGroup contains
long strings.
2. RowAggregation::initialize() allocates a memory buffer
for a NULL row. The size of this buffer is equal to the
row size for the output RowGroup. For the above scenario,
using the default group_concat_max_len (which is a server
variable that sets the maximum length of the GROUP_CONCAT string)
value of 1mb, the buffer size would be
(1mb + 64kb + some additional metadata). If the user sets
group_concat_max_len to a higher value, say 3gb, this buffer
size would be ~3gb. Now if the runtime initiates several
instances of RowAggregation, total memory consumption by
PrimProc could exceed the hardware memory limits causing the
OS OOM to kill the process. We fix this problem by again
enabling the StringStore for the NULL row allocation.
3. In the plugin code in buildAggregateColumn(), there is
an integer overflow when the server group_concat_max_len
variable (which is an uint32_t) is set to a value > INT32_MAX
(such as 3gb) and is assigned to
CalpontSystemCatalog::ColType::colWidth (which is an int32_t).
As a short term fix, we saturate the assigned value to colWidth
to INT32_MAX. Proper fix would be to upgrade
CalpontSystemCatalog::ColType::colWidth to an uint32_t.
This patch improves handling of NULLs in textual fields in ColumnStore.
Previously empty strings were considered NULLs and it could be a problem
if data scheme allows for empty strings. It was also one of major
reasons of behavior difference between ColumnStore and other engines in
MariaDB family.
Also, this patch fixes some other bugs and incorrect behavior, for
example, incorrect comparison for "column <= ''" which evaluates to
constant True for all purposes before this patch.
The idea is relatively simple - encode prefixes of collated strings as
integers and use them to compute extents' ranges. Then we can eliminate
extents with strings.
The actual patch does have all the code there but miss one important
step: we do not keep collation index, we keep charset index. Because of
this, some of the tests in the bugfix suite fail and thus main
functionality is turned off.
The reason of this patch to be put into PR at all is that it contains
changes that made CHAR/VARCHAR columns unsigned. This change is needed in
vectorization work.
* Fix clang warnings
* Remove vim tab guides
* initialize variables
* 'strncpy' output truncated before terminating nul copying as many bytes from a string as its length
* Fix ISO C++17 does not allow 'register' storage class specifier for outdated bison
* chars are unsigned on ARM, having if (ival < 0) always false
* chars are unsigned by default on ARM and comparison with -1 if always true
This patch fixes:
- MCOL-4614 calShowPartitions() precision loss for huge narrow decimal
- MCOL-4615 GROUP_CONCAT() precision loss for huge narrow decimal
- MCOL-4660 Narow decimal to string conversion is inconsistent about zero integral
Changes:
- Implementing Row::getDecimalField()
- Removing double arithmetic from the code printing DECIMAL values
in TypeHandlerXDecimal::format64() and GroupConcator::outputRow().
Using Decimal::toString() instead.
- Rewriting Decimal::toStringTSInt64(). The old implementation
was wrong, too complex and slow (used unnecessary memmove, memcpy).
An additional cleanup:
- Removing the ENGINE=COLUMNSTORE clause from tests for MCOL-4532 and MCOL-4640
type_decimal.test is combinations-aware. It's run two times with
default_storage_engine=MyISAM and default_storage_engine=COLUMNSTORE.
So the CREATE TABLE statements should not specify the engine explicitly.
- Adding --disable_warnings in the old fixed test.
We needed to suppress warnings when the MyISAM combination is being run.
Previously the table was erroneously created with ENGINE=COLUMNSTORE
even with the MyISAM combination run. So warning were not generated.
MCOL-4409 This patch combines VDecimal and Decimal and makes
IDB_Decimal an alias for the result class
MCOL-4409 More boilerplate reduction in Func_mod
Removed couple TSInt128::toType() methods
Rename packages to MariaDB-columnstore-engine, MariaDB-columnstore-libs
and MariaDB-columnstore-platform.
Also add the "columnstore-" prefix the the components so that MariaDB's
packaging system understands then and add a line to include them in
MariaDB's packaging.
In addition
* Fix S3 building for dist source build
* Fix Debian 10 dependency issue
* Fix git handling for dist builds
* Add support for MariaDB's RPM building
* Use MariaDB's PCRE and readline
* Removes a few dead files
* Fix Boost noncopyable includes
ExeMgr uses ResourceManager to count memory usage. If a usage exceeded
error occurs the counting wasn't reset and subsequent usage attempts in
the same ExeMgr thread would error.
This patch moves the in-class accounting for GroupConcat and others so
that it happens before the error is detected. The memory usage counter
is then decremented correctly on the class destructor.
pDictionaryScan won't work for BLOB/TEXT since it requires searching the
data file and rebuilding the token from matches. The tokens can't be
rebuild correctly due the bits in the token used for block counts. This
patch forces the use of pDictionaryStep instead for WHERE conditions.
In addition this patch adds support for TEXT/BLOB in various parts of
the job step processing. This fixes things like error 202 during an
UPDATE with a join condition on TEXT/BLOB columns.
This does the following:
* Switch resource manager to a singleton which reduces the amount of
times the XML data is scanned and objects allocated.
* Make the I_S tables use the FE implementation of the system catalog
* Make the I_S.columnstore_columns table use the RID list cache
* Make the extentmap pre-allocate a vector instead of many small allocs
