Skip to content

Instantly share code, notes, and snippets.

@jberkus
Last active August 29, 2015 14:14
Show Gist options
  • Select an option

  • Save jberkus/2d614fa3eccbad83d008 to your computer and use it in GitHub Desktop.

Select an option

Save jberkus/2d614fa3eccbad83d008 to your computer and use it in GitHub Desktop.
EC2 sizing for benchmarks.
Amazon EC2
Small Instance
m3.medium
3.5GB RAM
1 core
EBS storage, 1000 IOPS
R/W in-memory test
pgbench -i -s 100 --foreign-keys
pgbench -c 4 -T 900
R-O in-memory test
pgbench -i -s 100 --foreign-keys
pgbench -c 4 -T 900 -S
R/W on-disk test
pgbench -i -s 400 --foreign-keys
pgbench -c 4 -T 900
Tuned instance pg.conf:
shared_buffers = 1GB
wal_buffers = 32MB
effective_cache_size = 3GB
maintenance_work_mem = 128MB
random_page_cost = 1.5
wal_level = hot_standby
stats_temp_directory = '/pgramdisk/pg_stat_tmp'
checkpoint_segments = 32
checkpoint_completion_target = 0.9
Large Instance
r3.2xLarge
61GB RAM
8 cores
EBS storage, 5000 IOPS
xfs filesystem (if possible)
other HW/OS configuration tweaks
R/W in-memory test
pgbench -i -s 1000 --foreign-keys
pgbench -j 4 -c 16 -T 900
R-O in-memory test
pgbench -i -s 1000 --foreign-keys
pgbench -j 4 -c 16 -T 900 -S
R/W on-disk test
pgbench -i -s 7000 --foreign-keys
pgbench -j 4 -c 16 -T 900
Tuned instance pg.conf:
shared_buffers = 8GB
wal_buffers = 32MB
effective_cache_size = 50GB
maintenance_work_mem = 1GB
random_page_cost = 1.5
wal_level = hot_standby
stats_temp_directory = '/pgramdisk/pg_stat_tmp'
checkpoint_segments = 128
checkpoint_completion_target = 0.9
Large Instance with SSD:
r3.2xLarge
61GB RAM
8 cores
Internal storage
xfs filesystem (if possible)
other HW/OS configuration tweaks
Same parameters as above.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment