Open RiskyFeryansyahP opened 5 months ago
Hi @hibiken, if you agree with my proposal, I would be happy to implement it π
Hi @kamikazechaser, if you have time, please check my proposal request π
Any update guys?
Is there any update for this @hibiken?
Your issue seems fit to me, I also had this when I had a thousand jobs
Is your feature request related to a problem? Please describe. Currently, in our production environment, I have 9,000 scheduled entries that run once every week and once every month. However, I've noticed that the CPU utilization on my Redis spikes continuously. You can see this in the image below.
When I investigated the reason for these CPU spikes, I disabled the scheduler in
asynq
, and the CPU usage returned to normal. While examining the code for the scheduler inasynq
, I found that the heartbeater insideasynq
runs every 5 seconds (hardcoded) using a ticker. This process involves scanning through cron entries and performing WriteSchedulerEntries.Describe the solution you'd like I have an idea to configure the ticker duration based on
SchedulerOpts
. This way, the ticker duration is not hardcoded and can be defined through configuration when initializingNewPeriodicTaskManager
.Describe alternatives you've considered A clear and concise description of any alternative solutions or features you've considered.
Additional context Add any other context or screenshots about the feature request here.