How to Improve Azure SQL Database Availability on a Budget?

0
5
Asked By CuriousCat92 On

I'm currently using an Azure SQL Database on the Standard tier with 20 DTUs, costing about €25 per month. This works great for my performance needs since I expect the workload will stay below 100 DTUs for a while. However, I'm facing issues with availability. I need a reliable database but don't want to blow my budget while improving this.

So far, I've looked into several options:

1. **Failover Groups (Geo-Replication)**: It could cost around €50 per month for two Standard-tier instances, but I'd need to manage my own failover logic, which concerns me about the added complexity.

2. **Premium Tier (DTU Model)**: This would jump to about €400/month, which is not feasible for me.

3. **Hyperscale (vCore Model)**: This offers high availability as well but starts at around €500/month, also beyond my budget.

Essentially, I'm hoping to find a middle-ground solution that boosts availability without pushing costs significantly above €200/month. Is using failover groups and custom failover logic the best route, or is there something else I'm not considering?

2 Answers

Answered By CloudNinja44 On

I get where you’re coming from, but in all honesty, 30 minutes of downtime a month isn’t that bad, and it seems you’re still within the SLA of your current tier. Failover groups might be overkill for your needs and could end up causing more costs and complications than they're worth.

If SQL isn’t a must for your application, there might be more flexible, budget-friendly solutions out there. You could explore alternative databases or configurations that offer built-in redundancy without the leap in pricing. Make sure to weigh the trade-offs of what you really need versus what you’re willing to manage.

Answered By TechSavvy88 On

It sounds like you're having network-related issues since you mentioned the 'A transport-level error has occurred' message. That can pop up occasionally, but if it’s just about once a month, your situation isn't too dire.

Regarding failover groups, I think they might complicate things for you since you'd have to do manual failovers when alerted to a non-responsive database. If you're set on this, you'd need a solid strategy to make it feasible without causing more headaches. You might want to consider monitoring tools or service level agreements (SLAs) that naturally cover some downtime.

Alternatively, if you don’t want to manage the complexities of failover logic, maybe stick with the Standard tier and focus on optimizing the queries and connection strategies in your application instead.

Related Questions

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.