How to reduce cloud server cost
We reduced our cloud server cost by more than half to $210. We did by just following these simple steps:
Understand overall costs
Understand compute costs
You should check the types and sizes of virtual machines you’re running. Prices are different by region, on-demand requirement and the use of the fixed instance or environment.
Use real-time monitoring and analytics
Managed databases
If you are using a managed database, you need to understand the cost based on usage or per-hour rates.
Costs of storage
Understand the storage types you use, including data retrieval and data transfer fees. Optimizing this can drastically reduce storage costs.
领英推荐
Support costs
Do you really need support? I ended up never having to use free credits. Remember to unsubscribe from your free credit plan. It has to be done separately from your cloud plan or you will be charged a fixed amount monthly or yearly depending on your plan.
Bandwidth costs
These are basically data egress fees - check this cost if your business is bandwidth-sensitive.
Choose your cloud provider wisely
We do not prefer shared instances. We go for independent fixed empty boxes - cost is a bit higher, but affordable with manageable traffic
Locate idle cloud resources
We paid $300+ for months for backup of staging servers, database which were not in use. Finally closed it and saved a few bucks there. Similarly, check for the instances with minimum activity over a period of time. If not required, delete it.
Have different configurations
Experiment with different sizes, for both website and database. Understand your workloads and find the optimal configuration for your requirements.
Implement load balancing
Use if required. This will help in distributing workload across resources, preventing overuse and underuse.
Have a multi-cloud strategy
Every cloud service has its strengths and weaknesses. This will also support risk management and prevent vendor lock-in.
Nice work getting the cost down Jwalant Patel. You mention you experimented with different sizes for both website and database - how many attempts did it take to find the right VM or resource? Any tips for navigating this experimentation process?