[BUG] use alluxio to accelerate data, but it seems that it does not accelerate。
What is your environment(Kubernetes version, Fluid version, etc.)
alluixo controller image: fluidcloudnative/alluxioruntime-controller:v1.0.1-14eda3b runtime image: alluxio/alluxio-dev:2.9.0 fuse image: alluxio/alluxio-dev:2.9.0
Describe the bug
use alluxio to accelerate data, but it seems that it does not accelerate。
配置:1个worker, 使用ssd作为缓存介质,缓存容量大小设置为10G
1GiB 文件
业务pod和缓存worker pod不在同一节点上,读取远程缓存数据
社区版读文件, 执行3次:
22.1MB/s,
23.1MB/s,
25.4MB/s
直接挂载zos 1GiB 文件
业务pod直接挂底层zos
读文件, 执行3次:
19.8MB/s,
20.2MB/s,
22.5MB/s
测试语句
读文件
fio -filename=/data/pytorch/1g.txt -direct=1 -iodepth 64 -thread -rw=randread -ioengine=libaio -bs=128K -size=1G -numjobs=10 -runtime=180 -group_reporting -name=rand_1Gread_128K
@ruan875417 可以提供下AlluxioRuntime的配置吗?是否使用了本地磁盘作为缓存介质,远程数据源与Alluxio集群之间的带宽是多少?