Collector Worker Installation Customization
When you install the NorthStar application, a default number of collector workers are installed on the NorthStar server, depending on the number of cores in the CPU. This is regulated in order to optimize server resources, but you can change the number by using a provided script. Each installed worker starts a number of celery processes equal to the number of cores in the CPU plus one.
Table 1 describes the default number of workers installed according to the number of cores in the CPU.
Table 1: Default Worker Groups and Processes by Number of CPU Cores
Worker Groups Installed
Total Worker Processes
Minimum RAM Required
(CPUs +1) x 4 = 20
(CPUs +1) x 2 = 18
(CPUs +1) x 1 = 17
(CPUs +1) x 1 = 33
Use the config_celery_workers.sh script to change the number of worker groups installed (post-initial installation). You might want to make a change if, for example:
You upgrade your hardware with additional CPU cores and you want to increase the worker groups based on the new total number of cores.
You want to manually determine the number of workers to be started rather than using the automatically-applied formula.
NorthStar Controller System Requirements provides some guidance about memory requirements for various server uses and sizes.
You can also use the config_celery_workers.sh script to change the number of slave workers installed on a slave collector server. See Slave Collector Installation for Distributed Data Collection for more information about distributed data collection.
To change the number of worker groups installed, launch the config_celery_workers.sh script:
The available options are:
This option automatically determines the number of cores and calculates the number of worker groups to add accordingly, per the formulas in Table 1.
This option adds the specified number of worker groups. The following example starts six worker groups:
/opt/northstar/snmp-collector/scripts/config_celery_workers.sh -w 6