#What's this

mongo_bench.sh

You can get the benchmark of the local MongoDB

How to use

1. Setup local MongoDB in /usr/local/mongo.

   This script is requiring the PID file at /usr/local/mongo/logs/mongod.pid

   You can also use mongod.conf

2. Build mongo_bench.

   type "make" at this directory.

   See "How to build" for details

3. Run mongo_bench_mini.sh

   This bench is treating 400 MB data.

   First of all, you should run mongo_bench_mini.sh for test.

4. Run mongo_bench.sh

   This bench is treating 40 GB data.

   But it will continue to run more than a overnight.

#Ease to try this

make prepare
make test
./mongo_bench_mini.sh

#Outputs

####[insert]

Below outputs means,

Inserted 1000,000 around 400 bytes documents from 10 threads.

• It tooks 47.8652 seconds.
• mongod used 124.872 CPU in avarage.
• mongod could store 20892 documents per seconds.
• Client sends 385000000 bytes data.
• Actually,mongod saved 408000000 bytes data.
• Throuput was 7.67081 MB per seconds.

[output]

=== insert_test* : 400 : 1000000 : 10 ===
TIME: 47.8652 CPU: 124.872
 ALL>     20892 n/s, COL:  408000000 , TOTAL:  385000000 B,  0.358559 GB,   7.67081 MB/s,

####[update]

Below outputs means,

Updated number column from double to double.

• mongod updated 21000000 bytes in total.

[output]

=== update_test<double>* : 400 : 1000000 : 1 ===
TIME: 9.6226 CPU: 88.5416
 ALL>    103922 n/s, TOTAL:    21000000 B,2.08126e-06 MB/s,

####[update]

Below outputs means,

Updated number column from double (64bit) to long long (64bit).

[output]

=== update_test<long long>* : 400 : 1000000 : 1 ===
TIME: 34.6098 CPU: 96.1578
 ALL>   28893.6 n/s, TOTAL:   21000000 B,  0.578656 MB/s,

####[fetch by $in] Below outputs means,

Fetched 40000 documents by using $in operator from collection was housing 1000000 documents from 1 thread.

It means RANDOM fetch.

• Client received 15560000 bytes of BSON objects in total.

[output]

=== query_in_test* : 400 : 1000000 : 1 ===
TIME: 0.366799 CPU: 27.2629
 ALL>   40000,     109052 n/s, TOTAL:   15560000 B, 0.0144914 GB,   40.4559 MB/s,

Below outputs means, Fetched 40000 documents from collection was housing 1000000 documents from 100 thread.

=== query_in_test* : 400 : 1000000 : 100 ===

####[fetch by direct value]

Fetched 40000 documents by specifing value from collection was housing 1000000 documents from 1 thread.

It means RANDOM fetch.

[output]

=== query_test* : 400 : 1000000 : 1 ===

####[fetch by range]

Below outputs means,

Fetched 40000 documents by using $gte and $lt operators from collection was housing 1000000 documents from 1 thread.

It means SEQUENCIAL fetch.

[output]

=== query_range_test* : 400 : 1000000 : 1 ===
TIME: 0.092688 CPU: 86.3111
 ALL>   40000,     431555 n/s, TOTAL:   15560000 B, 0.0144914 GB,   160.098 MB/s,

#How to setup mongodb

sudo mkdir -p $@
sudo chmod 777 $@
mkdir -p $@/conf
mkdir -p $@/logs
mkdir -p $@/data
cp mongoctrl  $@/
cp mongod.conf $@/conf
wget http://fastdl.mongodb.org/linux/mongodb-linux-x86_64-2.4.3.tgz
tar xzf mongodb-linux-x86_64-2.4.3.tgz
rm mongodb-linux-x86_64-${MONGO_VERSION}.tgz
mv mongodb-linux-x86_64-2.4.3/bin $@/

#How to build

sudo yum install boost-devel.x86_64
export MONGO_VERSION=2.4.3
wget http://downloads.mongodb.org/cxx-driver/mongodb-linux-x86_64-${MONGO_VERSION}.tgz
tar xzf mongodb-linux-x86_64-${MONGO_VERSION}.tgz
pushd mongo-cxx-driver-v2.4
scons
popd

#How to run

g++ mongo_bench.cpp -o mongo_bench -O3 -pthread  -L mongo-cxx-driver-v2.4/build/ -l mongoclient -lboost_thread-mt -lboost_filesystem -I mongo-cxx-driver-v2.4/src/
nice ./mongo_bench 127.0.0.1:27017 test.TEST `cat /usr/local/mongo/logs/mongod.pid` 4000 100000 10