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
When I try FEMU Black box SSD, the write throughput is always around 1180MB/s.
I use default FEMU configuration and FIO tool to measure raw block device. Although I change fio configuration to sequential writes (128KB) and random writes (4KB) multiple times, the throughput remain same (1180MB/s). And I when continue to randomly write entire device, I didn't see any performance degradation as real NAND SSDs.
I'm wondering how can I reproduce GC influence with FEMU? The Black box SSD is expected to show lower write throughput when GC is triggered.
Thanks
The text was updated successfully, but these errors were encountered:
When I enable FEMU debug log, I find GC is indeed triggered. It seems like GC has visible influence on tail latency, but it hasn't too much visible influence on throughput (around 1-2% throughput degradation for writes).
Hi Huaicheng,
When I try FEMU Black box SSD, the write throughput is always around 1180MB/s.
I use default FEMU configuration and FIO tool to measure raw block device. Although I change fio configuration to sequential writes (128KB) and random writes (4KB) multiple times, the throughput remain same (1180MB/s). And I when continue to randomly write entire device, I didn't see any performance degradation as real NAND SSDs.
I'm wondering how can I reproduce GC influence with FEMU? The Black box SSD is expected to show lower write throughput when GC is triggered.
Thanks
The text was updated successfully, but these errors were encountered: