To prevent large memory allocations.
This series shows over 3% improvement in perf-simple-query throughput.
```
$ build/release/scylla perf-simple-query --default-log-level=error --smp=1 --random-seed=1855519715
random-seed=1855519715
enable-cache=1
Running test with config: {partitions=10000, concurrency=100, mode=read, query_single_key=no, counters=no}
Disabling auto compaction
Creating 10000 partitions...
Before:
random-seed=1775976514
enable-cache=1
enable-index-cache=1
sstable-summary-ratio=0.0005
sstable-format=me
Running test with config: {partitions=10000, concurrency=100, mode=read, query_single_key=no, counters=no}
Disabling auto compaction
Creating 10000 partitions...
336345.11 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32788 insns/op, 12430 cycles/op, 0 errors)
348748.14 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32794 insns/op, 12335 cycles/op, 0 errors)
349012.63 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32800 insns/op, 12326 cycles/op, 0 errors)
350629.97 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32770 insns/op, 12270 cycles/op, 0 errors)
348585.00 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32804 insns/op, 12338 cycles/op, 0 errors)
throughput:
mean= 346664.17 standard-deviation=5825.77
median= 348748.14 median-absolute-deviation=2348.46
maximum=350629.97 minimum=336345.11
instructions_per_op:
mean= 32791.35 standard-deviation=13.60
median= 32794.47 median-absolute-deviation=8.65
maximum=32804.45 minimum=32769.57
cpu_cycles_per_op:
mean= 12340.05 standard-deviation=57.57
median= 12335.05 median-absolute-deviation=13.94
maximum=12430.42 minimum=12270.28
After:
random-seed=1775976514
enable-cache=1
enable-index-cache=1
sstable-summary-ratio=0.0005
sstable-format=me
Running test with config: {partitions=10000, concurrency=100, mode=read, query_single_key=no, counters=no}
Disabling auto compaction
Creating 10000 partitions...
353770.85 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32762 insns/op, 11893 cycles/op, 0 errors)
364447.98 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32738 insns/op, 11818 cycles/op, 0 errors)
365268.97 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32734 insns/op, 11788 cycles/op, 0 errors)
344304.87 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32746 insns/op, 12506 cycles/op, 0 errors)
362263.57 tps ( 58.1 allocs/op, 0.0 logallocs/op, 14.1 tasks/op, 32756 insns/op, 11888 cycles/op, 0 errors)
throughput:
mean= 358011.25 standard-deviation=8916.76
median= 362263.57 median-absolute-deviation=6436.74
maximum=365268.97 minimum=344304.87
instructions_per_op:
mean= 32747.06 standard-deviation=11.85
median= 32745.80 median-absolute-deviation=9.36
maximum=32762.18 minimum=32734.01
cpu_cycles_per_op:
mean= 11978.65 standard-deviation=298.06
median= 11887.96 median-absolute-deviation=160.96
maximum=12505.72 minimum=11788.49
```
Refs #28511
(Refs rather than Fixes for the lack of a reproducer unit test)
* No backport needed as the issue is rare and not severe
Closes scylladb/scylladb#28631
* github.com:scylladb/scylladb:
query: result_set: change row member to a chunked vector
query: result_set_row: make noexcept
query: non_null_data_value: assert is_nothrow_move_constructible and assignable
types: data_value: assert is_nothrow_move_constructible and assignable
Scylla in-source tests.
For details on how to run the tests, see docs/dev/testing.md
Shared C++ utils, libraries are in lib/, for Python - pylib/
alternator - Python tests which connect to a single server and use the DynamoDB API unit, boost, raft - unit tests in C++ cqlpy - Python tests which connect to a single server and use CQL topology* - tests that set up clusters and add/remove nodes cql - approval tests that use CQL and pre-recorded output rest_api - tests for Scylla REST API Port 9000 scylla-gdb - tests for scylla-gdb.py helper script nodetool - tests for C++ implementation of nodetool
If you can use an existing folder, consider adding your test to it. New folders should be used for new large categories/subsystems, or when the test environment is significantly different from some existing suite, e.g. you plan to start scylladb with different configuration, and you intend to add many tests and would like them to reuse an existing Scylla cluster (clusters can be reused for tests within the same folder).
To add a new folder, create a new directory, and then
copy & edit its suite.ini.