in PVE 4.1, for LXC to mount host file system, it need to add config
However, I found it limits @ mp9 and does not process mp10 and forth
Another huge performance issue I found out that some how zfs performance seemed to be capped!
I ran following command on Dual Intel Xeon with 64 GB ECC RAM with NVMe SSD as log and cache using PVE 4.1 and I got following result
Same test ran on a i7 3770t CPU with 16 GB None ECC RAM without any SSD as log and cache use Ubuntu 15.10, and I got floowing result
This does not make sense at all.
Code:
mp0: /target/test,mp=/target
However, I found it limits @ mp9 and does not process mp10 and forth
Another huge performance issue I found out that some how zfs performance seemed to be capped!
I ran following command on Dual Intel Xeon with 64 GB ECC RAM with NVMe SSD as log and cache using PVE 4.1 and I got following result
Code:
root@nas:~# dd if=/dev/zero of=/rpool/output bs=4k count=16000k; rm -f /rpool/ou
tput
16384000+0 records in
16384000+0 records out
67108864000 bytes (67 GB) copied, 216.258 s, 310 MB/s
Same test ran on a i7 3770t CPU with 16 GB None ECC RAM without any SSD as log and cache use Ubuntu 15.10, and I got floowing result
Code:
root@tmedia:~# dd if=/dev/zero of=/rpool/output bs=4k count=16000k; rm -f /rpool/output
16384000+0 records in
16384000+0 records out
67108864000 bytes (67 GB) copied, 113.787 s, 590 MB/s
This does not make sense at all.