wa impact

lelunicu · 09-25-2019, 09:20 AM

hi,
this VM (with 8 vcpu)is running rhel and top
%Cpu(s): 65.3 us, 13.4 sy, 0.0 ni, 11.4 id, 8.9 wa

So wa is 8.9.

iostat -d 2 3
device tps
xvdg 2263.76
xvdj 1103.34
xvdh 439.76
xvdi 31.80

free -mh
total used free shared buff/cache available
83G 69G 6.8G 4.7G 7.1G 7.3G
Swap:15G 389M 15G

xvdg and xvdj handle the most i/o.
how can be reduced the wa?
Being a VM we can not calculate the real IOPS.right?
tnks

scasey · 09-25-2019, 09:26 AM

Quote:

Originally Posted by lelunicu

hi,
this VM (with 8 vcpu)is running rhel and top
%Cpu(s): 65.3 us, 13.4 sy, 0.0 ni, 11.4 id, 8.9 wa

So wa is 8.9.

iostat -d 2 3
device tps
xvdg 2263.76
xvdj 1103.34
xvdh 439.76
xvdi 31.80

free -mh
total used free shared buff/cache available
83G 69G 6.8G 4.7G 7.1G 7.3G
Swap:15G 389M 15G

xvdg and xvdj handle the most i/o.
how can be reduced the wa?
Being a VM we can not calculate the real IOPS.right?
tnks

What are you talking about? (Sheesh)
If RHEL, then open a ticket with RedHat...or redo your OP so that it makes some sense.
...and use code tags when posting output.

berndbausch · 09-25-2019, 09:39 AM

Quote:

Originally Posted by lelunicu

hi,
this VM (with 8 vcpu)is running rhel and top
%Cpu(s): 65.3 us, 13.4 sy, 0.0 ni, 11.4 id, 8.9 wa

So wa is 8.9.

iostat -d 2 3
device tps
xvdg 2263.76
xvdj 1103.34
xvdh 439.76
xvdi 31.80

free -mh
total used free shared buff/cache available
83G 69G 6.8G 4.7G 7.1G 7.3G
Swap:15G 389M 15G

xvdg and xvdj handle the most i/o.
how can be reduced the wa?
Being a VM we can not calculate the real IOPS.right?
tnks

You reduce the wait time by reducing disk I/O. But why is wa a problem?

I don't know what you mean by "real IOPS".