* Convert hashSets in parallel before merge
Before merge, if one of the lhs and rhs is singleLevelSet and the other is twoLevelSet,
then the SingleLevelSet will call convertToTwoLevel(). The convert process is not in parallel
and it will cost lots of cycle if it cosume all the singleLevelSet.
The idea of the patch is to convert all the singleLevelSets to twoLevelSets in parallel if
the hashsets are not all singleLevel or not all twoLevel.
I have tested the patch on Intel 2 x 112 vCPUs SPR server with clickbench and latest upstream
ClickHouse.
Q5 has got a big 264% performance improvement and 24 queries have got at least 5% performance
gain. The overall geomean of 43 queries has gained 7.4% more than the base code.
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
* add resize() for the data_vec in parallelizeMergePrepare()
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
* Add the performance test prepare_hash_before_merge.xml
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
* Fit the CI to rename the data set from hits_v1 to test.hits.
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
* remove the redundant branch in UniqExactSet
Co-authored-by: Nikita Taranov <nickita.taranov@gmail.com>
* Remove the empty methods and add throw exception in parallelizeMergePrepare()
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
---------
Signed-off-by: Jiebin Sun <jiebin.sun@intel.com>
Co-authored-by: Nikita Taranov <nickita.taranov@gmail.com>
Adding more processors for parallelize_output_from_storages is not a
costless operation (I've experienced some issues in production because
of this), and it is not easy to fix in a normal way, so let's disable it
for now.
Before this patch:
- INSERT INTO input SELECT * FROM numbers(10e6) SETTINGS parallelize_output_from_storages=1, min_insert_block_size_rows=1000
0 rows in set. Elapsed: 3.648 sec. Processed 20.00 million rows, 120.00 MB (5.48 million rows/s., 32.90 MB/s.)
- INSERT INTO input SELECT * FROM numbers(10e6) SETTINGS parallelize_output_from_storages=0, min_insert_block_size_rows=1000
0 rows in set. Elapsed: 1.851 sec. Processed 20.00 million rows, 120.00 MB (10.80 million rows/s., 64.82 MB/s.)
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
As it turns out, HashMap/PackedHashMap works great even with max load
factor of 0.99. By "great" I mean it least it works faster then
google sparsehash, and not to mention it's friendliness to the memory
allocator (it has zero fragmentation since it works with a continuious
memory region, in comparison to the sparsehash that doing lots of
realloc, which jemalloc does not like, due to it's slabs).
Here is a table of different setups:
settings | load (sec) | read (sec) | read (million rows/s) | bytes_allocated | RSS
- | - | - | - | - | -
HASHED upstream | - | - | - | - | 35GiB
SPARSE_HASHED upstream | - | - | - | - | 26GiB
- | - | - | - | - | -
sparse_hash_map glibc hashbench | - | - | - | - | 17.5GiB
sparse_hash_map packed allocator | 101.878 | 231.48 | 4.32 | - | 17.7GiB
PackedHashMap 0.5 | 15.514 | 42.35 | 23.61 | 20GiB | 22GiB
hashed 0.95 | 34.903 | 115.615 | 8.65 | 16GiB | 18.7GiB
**PackedHashMap 0.95** | **93.6** | **19.883** | **10.68** | **10GiB** | **12.8GiB**
PackedHashMap 0.99 | 26.113 | 83.6 | 11.96 | 10GiB | 12.3GiB
As it shows, PackedHashMap with 0.95 max_load_factor, eats 2.6x less
memory then SPARSE_HASHED in upstream, and it also 2x faster for read!
v2: fix grower
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
In case you want dictionary optimized for memory, SPARSE_HASHED is not
always gives you what you need.
Consider the following example <UInt64, UInt16> as <Key, Value>, but
this pair will also have a 6 byte padding (on amd64), so this is almost
40% of space wastage.
And because of this padding, even google::sparse_hash_map, does not make
picture better, in fact, sparse_hash_map is not very friendly to memory
allocators (especially jemalloc).
Here are some numbers for dictionary with 1e9 elements and UInt64 as
key, and UInt16 as value:
settings | load (sec) | read (sec) | read (million rows/s) | bytes_allocated | RSS
HASHED upstream | - | - | - | - | 35GiB
SPARSE_HASHED upstream | - | - | - | - | 26GiB
- | - | - | - | - | -
sparse_hash_map glibc hashbench | - | - | - | - | 17.5GiB
sparse_hash_map packed allocator | 101.878 | 231.48 | 4.32 | - | 17.7GiB
PackedHashMap | 15.514 | 42.35 | 23.61 | 20GiB | 22GiB
As you can see PackedHashMap looks way more better then HASHED, and
even better then SPARSE_HASHED, but slightly worse then sparse_hash_map
with packed allocator (it is done with a custom patch to google
sparse_hash_map).
v2: rebase on top of bucket_count fix
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
