benchmark-cars has been rewritten to perform WAY FASTER benchmarks.

This commit is contained in:
Philippe PITTOLI 2024-05-12 14:04:22 +02:00
parent b3269acdad
commit f43669b3b7
2 changed files with 96 additions and 80 deletions

View File

@ -6,12 +6,12 @@ require "./db-cars.cr"
# ENV["CARNAME"] rescue "Corvet-#{(db_size/2).to_i}" # ENV["CARNAME"] rescue "Corvet-#{(db_size/2).to_i}"
# ENV["CARCOLOR"] rescue "red" # ENV["CARCOLOR"] rescue "red"
# ENV["CARKEYWORD"] rescue "spacious" # ENV["CARKEYWORD"] rescue "spacious"
# ENV["DBSIZE"] rescue 40_000 # ENV["DBSIZE"] rescue 50_000
# ENV["DBSIZE_START"] rescue 5_000 # ENV["DBSIZE_START"] rescue 1_000
# ENV["DBSIZE_INCREMENT"] rescue 1_000 # ENV["DBSIZE_INCREMENT"] rescue 1_000
# ENV["REPORT_DIR"] rescue "results" # ENV["REPORT_DIR"] rescue "results"
# ENV["NBRUN"] rescue 100 # ENV["NBRUN"] rescue 100
# ENV["INDEXESMAXIT"] rescue 5000 # ENV["MAXINDEXES"] rescue 5_000
class DODB::Storage(V) class DODB::Storage(V)
def empty_db def empty_db
@ -22,32 +22,58 @@ end
class Context class Context
class_property report_dir = "results" class_property report_dir = "results"
class_property max_indexes = 5_000
class_property nb_run = 100
class_property from = 1_000
class_property to = 50_000
class_property incr = 1_000
end end
def report(storage, nb_run, name, &block) def report(storage, name, &block)
durations = run_n_times nb_run, &block durations = run_n_times Context.nb_run, &block
File.open("#{Context.report_dir}/#{name}", "w") do |file| File.open("#{Context.report_dir}/#{name}.raw", "w") do |file|
durations.each do |d| durations.each do |d|
file.puts d file.puts d
end end
end end
avr = durations.reduce { |a, b| a + b } / nb_run avr = durations.reduce { |a, b| a + b } / Context.nb_run
puts "#{name}: #{avr}" puts "#{name}: #{avr}"
avr avr
end end
def prepare_env(storage, db_size, name, s_index, s_partition, s_tags, &) def long_operation(text)
STDOUT.write "add #{db_size} values to #{name}\r".to_slice STDOUT.write "#{text}\r".to_slice
index_max_iterations = ENV["INDEXESMAXIT"].to_i rescue 5_000 yield
add_cars storage, db_size, max_it_tags: (db_size > index_max_iterations ? 0 : 5000)
STDOUT.write " \r".to_slice STDOUT.write " \r".to_slice
yield storage, name, s_index, s_partition, s_tags
storage.rm_storage_dir
end end
def batch(db_size, nb_run) def verbose_add_cars(storage, nbcars, name, max_indexes)
long_operation "add #{nbcars} values to #{name}" do
add_cars storage, nbcars, max_indexes: max_indexes
end
end
# Add first entries, then loop: speed tests, add entries.
def prepare_env(storage, name, s_index, s_partition, s_tags, &)
verbose_add_cars storage, Context.from, name, max_indexes: Context.max_indexes
current = Context.from
to = Context.to
incr = Context.incr
while current < to
yield storage, current, name, s_index, s_partition, s_tags
break if current + incr >= to
verbose_add_cars storage, incr, name, max_indexes: Context.max_indexes
current += incr
end
long_operation "removing #{name} data" { storage.rm_storage_dir }
end
def batch()
cars_ram = DODB::RAMOnlySpecDataBase(Car).new cars_ram = DODB::RAMOnlySpecDataBase(Car).new
cars_cached = DODB::CachedSpecDataBase(Car).new cars_cached = DODB::CachedSpecDataBase(Car).new
cars_semi = DODB::SpecDataBase(Car).new "-semi" cars_semi = DODB::SpecDataBase(Car).new "-semi"
@ -58,42 +84,38 @@ def batch(db_size, nb_run)
semi_Sby_name, semi_Sby_color, semi_Sby_keywords = cached_indexes cars_semi semi_Sby_name, semi_Sby_color, semi_Sby_keywords = cached_indexes cars_semi
uncached_Sby_name, uncached_Sby_color, uncached_Sby_keywords = uncached_indexes cars_uncached uncached_Sby_name, uncached_Sby_color, uncached_Sby_keywords = uncached_indexes cars_uncached
car_name_to_search = ENV["CARNAME"] rescue "Corvet-#{(db_size/2).to_i}"
car_color_to_search = ENV["CARCOLOR"] rescue "red"
car_keyword_to_search = ENV["CARKEYWORD"] rescue "spacious"
puts ""
puts "NEW BATCH: db-size #{db_size}"
puts ""
puts "search by index '#{car_name_to_search}', partition '#{car_color_to_search}', tag '#{car_keyword_to_search}'"
fn = ->(storage : DODB::Storage(Car), fn = ->(storage : DODB::Storage(Car),
current_db_size : Int32,
name : String, name : String,
search_name : DODB::Index(Car), search_name : DODB::Index(Car),
search_color : DODB::Partition(Car), search_color : DODB::Partition(Car),
search_keywords : DODB::Tags(Car)) { search_keywords : DODB::Tags(Car)) {
report(storage, nb_run, "#{name}_#{db_size}_index") do
corvet = search_name.get car_name_to_search name_to_search = ENV["CARNAME"] rescue "Corvet-#{(current_db_size/2).to_i}"
color_to_search = ENV["CARCOLOR"] rescue "red"
keyword_to_search = ENV["CARKEYWORD"] rescue "spacious"
puts "NEW BATCH: db-size #{current_db_size}, name: '#{name_to_search}', color: '#{color_to_search}', tag: '#{keyword_to_search}'"
report(storage, "#{name}_#{current_db_size}_index") do
corvet = search_name.get name_to_search
end end
report(storage, nb_run, "#{name}_#{db_size}_partitions") do report(storage, "#{name}_#{current_db_size}_partitions") do
corvet = search_color.get? car_color_to_search corvet = search_color.get? color_to_search
end end
report(storage, nb_run, "#{name}_#{db_size}_tags") do report(storage, "#{name}_#{current_db_size}_tags") do
corvet = search_keywords.get? car_keyword_to_search corvet = search_keywords.get? keyword_to_search
end end
puts ""
} }
prepare_env cars_ram, db_size, "ram", ram_Sby_name, ram_Sby_color, ram_Sby_keywords, &fn prepare_env cars_cached, "cached", cached_Sby_name, cached_Sby_color, cached_Sby_keywords, &fn
prepare_env cars_cached, db_size, "cached", cached_Sby_name, cached_Sby_color, cached_Sby_keywords, &fn prepare_env cars_ram, "ram", ram_Sby_name, ram_Sby_color, ram_Sby_keywords, &fn
prepare_env cars_semi, db_size, "semi", semi_Sby_name, semi_Sby_color, semi_Sby_keywords, &fn prepare_env cars_semi, "semi", semi_Sby_name, semi_Sby_color, semi_Sby_keywords, &fn
prepare_env cars_uncached, db_size, "uncached", uncached_Sby_name, uncached_Sby_color, uncached_Sby_keywords, &fn prepare_env cars_uncached, "uncached", uncached_Sby_name, uncached_Sby_color, uncached_Sby_keywords, &fn
end end
def perform_add(storage : DODB::Storage(Car), nb_run) def perform_add(storage : DODB::Storage(Car))
corvet0 = Car.new "Corvet", "red", [ "shiny", "impressive", "fast", "elegant" ] corvet0 = Car.new "Corvet", "red", [ "shiny", "impressive", "fast", "elegant" ]
i = 0 i = 0
perform_benchmark_average nb_run, do perform_benchmark_average Context.nb_run, do
corvet = corvet0.clone corvet = corvet0.clone
corvet.name = "Corvet-#{i}" corvet.name = "Corvet-#{i}"
storage << corvet storage << corvet
@ -101,7 +123,7 @@ def perform_add(storage : DODB::Storage(Car), nb_run)
end end
end end
def batch_add(nb_run) def batch_add()
cars_ram = DODB::RAMOnlySpecDataBase(Car).new cars_ram = DODB::RAMOnlySpecDataBase(Car).new
cars_cached = DODB::CachedSpecDataBase(Car).new cars_cached = DODB::CachedSpecDataBase(Car).new
cars_semi = DODB::SpecDataBase(Car).new "-semi" cars_semi = DODB::SpecDataBase(Car).new "-semi"
@ -112,17 +134,17 @@ def batch_add(nb_run)
cached_indexes cars_semi cached_indexes cars_semi
uncached_indexes cars_uncached uncached_indexes cars_uncached
avr = perform_add(cars_ram, nb_run) avr = perform_add(cars_ram)
puts "(ram db and indexes) add a value (average on #{nb_run} tries): #{avr}" puts "(ram db and indexes) add a value (average on #{Context.nb_run} tries): #{avr}"
avr = perform_add(cars_cached, nb_run) avr = perform_add(cars_cached)
puts "(cached db and indexes) add a value (average on #{nb_run} tries): #{avr}" puts "(cached db and indexes) add a value (average on #{Context.nb_run} tries): #{avr}"
avr = perform_add(cars_semi, nb_run) avr = perform_add(cars_semi)
puts "(uncached db but cached indexes) add a value (average on #{nb_run} tries): #{avr}" puts "(uncached db but cached indexes) add a value (average on #{Context.nb_run} tries): #{avr}"
avr = perform_add(cars_uncached, nb_run) avr = perform_add(cars_uncached)
puts "(uncached db and indexes) add a value (average on #{nb_run} tries): #{avr}" puts "(uncached db and indexes) add a value (average on #{Context.nb_run} tries): #{avr}"
cars_ram.rm_storage_dir cars_ram.rm_storage_dir
cars_cached.rm_storage_dir cars_cached.rm_storage_dir
@ -130,32 +152,20 @@ def batch_add(nb_run)
cars_uncached.rm_storage_dir cars_uncached.rm_storage_dir
end end
ENV["REPORT_DIR"]?.try do |report_dir| ENV["REPORT_DIR"]?.try { |report_dir| Context.report_dir = report_dir }
Context.report_dir = report_dir
end
Dir.mkdir_p Context.report_dir Dir.mkdir_p Context.report_dir
nb_run = ENV["NBRUN"].to_i rescue 100 ENV["MAXINDEXES"]?.try { |it| Context.max_indexes = it.to_i }
db_size_max = ENV["DBSIZE"].to_i rescue 40_000 ENV["NBRUN"]?.try { |it| Context.nb_run = it.to_i }
current_size = ENV["DBSIZE_START"].to_i rescue 5_000 ENV["DBSIZE"]?.try { |it| Context.to = it.to_i }
size_increment = ENV["DBSIZE_INCREMENT"].to_i rescue 1_000 ENV["DBSIZE_START"]?.try { |it| Context.from = it.to_i }
index_max_iterations = ENV["INDEXESMAXIT"].to_i rescue 5_000 ENV["DBSIZE_INCREMENT"]?.try { |it| Context.incr = it.to_i }
# ENV["DBSIZE"] rescue 40_000 pp! Context.nb_run
# ENV["DBSIZE_START"] rescue 5_000 pp! Context.from
# ENV["DBSIZE_INCREMENT"] rescue 1_000 pp! Context.to
# ENV["REPORT_DIR"] rescue "results" pp! Context.incr
# ENV["NBRUN"] rescue 100 pp! Context.max_indexes
# ENV["INDEXESMAXIT"] rescue 5000
pp! nb_run batch
pp! db_size_max batch_add
pp! current_size
pp! size_increment
pp! index_max_iterations
while current_size <= db_size_max
batch current_size, nb_run
current_size += size_increment
end
batch_add nb_run

View File

@ -51,8 +51,12 @@ def uncached_indexes(storage : DODB::Storage)
return n, c, k return n, c, k
end end
def add_cars(storage : DODB::Storage, nb_iterations : Int32, from = 0, max_it_tags = 5000) # `max_indexes` limits the number of indexes (partitions and tags).
i = from # Once the last index (db last_index/5) is above this value, the following
# cars won't be tagged nor partitionned.
def add_cars(storage : DODB::Storage, nb_iterations : Int32, max_indexes = 5000)
last_index = ((storage.last_index + 1) / 5).to_i
i = 0
car1 = Car.new "Corvet", "red", [ "shiny", "impressive", "fast", "elegant" ] car1 = Car.new "Corvet", "red", [ "shiny", "impressive", "fast", "elegant" ]
car2 = Car.new "Bullet-GT", "blue", [ "shiny", "fast", "expensive" ] car2 = Car.new "Bullet-GT", "blue", [ "shiny", "fast", "expensive" ]
car3 = Car.new "Deudeuche", "beige", [ "curvy", "sublime" ] car3 = Car.new "Deudeuche", "beige", [ "curvy", "sublime" ]
@ -60,13 +64,15 @@ def add_cars(storage : DODB::Storage, nb_iterations : Int32, from = 0, max_it_ta
car5 = Car.new "C-MAX", "gray", [ "spacious", "affordable" ] car5 = Car.new "C-MAX", "gray", [ "spacious", "affordable" ]
while i < nb_iterations while i < nb_iterations
car1.name = "Corvet-#{i}" car1.name = "Corvet-#{last_index}"
car2.name = "Bullet-GT-#{i}" car2.name = "Bullet-GT-#{last_index}"
car3.name = "Deudeuche-#{i}" car3.name = "Deudeuche-#{last_index}"
car4.name = "Ford-5-#{i}" car4.name = "Ford-5-#{last_index}"
car5.name = "C-MAX-#{i}" car5.name = "C-MAX-#{last_index}"
if i >= max_it_tags last_index += 1
if last_index > max_indexes
car1.color = DODB.no_index car1.color = DODB.no_index
car2.color = DODB.no_index car2.color = DODB.no_index
car3.color = DODB.no_index car3.color = DODB.no_index