You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, the StoragePool layer manages the queue and thread pool and dispatches the requests to the Vdevs. This approach of maintaining a single queue, in certain scenarios, could hamper Haura from achieving parallelization at the Vdev level and could result in a bottleneck.
E.g.
Consider the following scenario where the queue size is 10, the thread pool size is 5, the first 9 requests belong to HDD and the last one belongs to SSD NVMe. In this case, requests for SSD NVMe cannot be processed until HDD requests are finished.
The text was updated successfully, but these errors were encountered:
Currently, the StoragePool layer manages the queue and thread pool and dispatches the requests to the Vdevs. This approach of maintaining a single queue, in certain scenarios, could hamper Haura from achieving parallelization at the Vdev level and could result in a bottleneck.
E.g.
Consider the following scenario where the queue size is 10, the thread pool size is 5, the first 9 requests belong to HDD and the last one belongs to SSD NVMe. In this case, requests for SSD NVMe cannot be processed until HDD requests are finished.
The text was updated successfully, but these errors were encountered: