Exam DP-300 topic 3 question 46 discussion

I am checking my SQL DB in West US: General purpose: 2 vCores - $260 - IOPS 640 - Log rate 9 MiB/sec 4 vCores - $480 - IOPS 1280 - Log rate 18 MiB/sec Business Critical: 2 vCores - $525 - IOPS 8000 - Log rate 24 MiB/sec 4 vCores - $970 - IOPS 16000 - Log rate 48 MiB/sec Since question says minimize costs and doesn't say to what extent we should reduce Log IO, I would pick "Increase the number of vCores" since $480 is less than $525... But otherwise it would make more sense to go Business Critical 2 vCores for $525 as it has far more IOPS and throughput to offer.

The correct answer is: C. Perform a checkpoint operation.

I belive it's A. When you increase number of vCores, you'll get higher resource limits incl. IOPS. https://learn.microsoft.com/en-us/azure/azure-sql/database/resource-limits-vcore-single-databases?view=azuresql It should be less expensive than promoting of service level.

it seems as if the only available option is A change service tire which is not cost-effective

A. Change Service tier to Business Critical. - seems the only solution. B. Increase the number of vCores. - relates to MI being in the Business Critical already. it is not (https://learn.microsoft.com/en-us/answers/questions/603214/log-io-metrics-is-hitting-100) C. Perform a checkpoint operation. - usually for shrinking log files, not related to log performance being poor. if anything, will worsen the situation by flushing data from memory to disk. the operation has to be logged as well. D. Change Recovery model to Simple. - logging transactions is not related to recovery models, transaction log management is (log backups or not). irrelevant for the issue at hand.

https://learn.microsoft.com/en-us/answers/questions/603214/log-io-metrics-is-hitting-100

since it is hitting max IOPS limitation and not storage space limits, upgrading to BC tier will increase max iops throughput . it won't obviously minimize the cost, but given the choices i think it is the only option

Both C and D can help to reduce the MAX Log IO performance. A checkpoint operation is a cost-effective way of managing transaction log growth in an Azure SQL database. It flushes the dirty pages from memory to disk and marks the transaction log as reusable, freeing up space in the log file. Changing the recovery model to Simple reduces log space usage, as it truncates the transaction log after every checkpoint. This can help address high MAX Log IO performance, but it also means that you will lose the ability to recover the database to a specific point in time. I think C is the best answer.