PostgreSQL Performance Tuning
If you are running ServiceDesk Plus MSP with bundled Postgres, you can benefit by allotting more resources for the Postgres database. For ServiceDesk Plus MSP 13 or later, you can use the configuration given below if you have medium (more than 100 thousand non-archived tickets) to a large workload.
In case you have a higher workload, you need to have better RAM and CPU capacity in the application server. These resources will be shared by the Postgres server and ServiceDesk Plus MSP's JVM (Java Virtual Machine). Consider the following split up.
| Customer Type | Physical RAM | PG | Java max heap size | CPU |
| Default | 4GB | 1GB | 2.5GB | Dual-core with hyperthreading / Quad-core |
| Medium | 8GB | 2GB | 4GB | Dual-core with hyperthreading / Quad-core |
| Large | 12GB | 4GB | 4GB | Quad-core with hyperthreading / Octa-core |
| Large | 16GB | 8GB | 6GB | Quad-core with hyperthreading / Octa-core |
Note: It is recommended to leave some RAM space for Operating System and other small applications running on the server machine. It is not recommended to run any other server / ManageEngine applications like AD plus, DC, etc., on the same machine.
Below is a sample configuration for a 2GB setup, modify the similar configuration accordingly in ServiceDesk/pgsql/ext_conf/postgres_ext.conf to suit your requirement.
# Uncomment below params when 2 GB RAM can be allotted for Postgres DB server
#
# shared_buffers = 512MB
# effective_cache_size = 2GB
# maintenance_work_mem = 128MB
# work_mem = 16MB
# wal_buffers = 16MB
#####
Uncomment below params when SSD is used
# random_page_cost = 1.1
#####