Hi all,

We are experiencing poor performance in one of Oracle environements, higher service time on Solaris 10 LDOMs. Oracle support suggests that it's the backend storage mis-configuration. I'd like to understand some entry point to troubleshoot this issue/s.

The DBAs and the application users suggest that the performance is sluggish at times of higher loads. This is a current snapshot and i have in the past experienced a performance degradation when the service times crept to double figures.

The disks are zfs vdevs

snip:

r/s    w/s   kr/s   kw/s wait actv wsvc_t asvc_t  %w  %b device

    4.0    5.4  129.1  115.7  0.0  0.0    0.0    2.8   0   2 c0d131

    0.0    0.0    0.1    0.0  0.0  0.0    0.0    1.3   0   0 c0d132

    0.0    0.0    0.1    0.0  0.0  0.0    0.0    0.4   0   0 c0d136

    5.7    5.6  573.7  483.0  0.0  0.2    0.2   15.6   0   6 c0d120

    5.7    5.6  575.4  483.1  0.0  0.2    0.2   15.8   0   6 c0d121

    5.7    5.6  575.1  483.1  0.0  0.2    0.2   16.2   0   6 c0d122

    0.1    0.0    4.9    0.1  0.0  0.0    0.0   28.2   0   0 c0d129

   56.2    3.7 6877.8  245.3  0.0  1.9    0.0   31.9   0  22 c0d137

   56.2    3.7 6875.1  245.2  0.0  1.9    0.0   32.0   0  22 c0d138

   56.2    3.7 6876.5  245.0  0.0  1.9    0.0   32.0   0  22 c0d139

    3.5    2.5  396.2  240.5  0.0  0.1    0.1   17.3   0   3 c0d123

The stats are similar in the control domain which is Running Solaris 11.

The disks have a VPID of : Product ID:           Invista.

These disks are virtualised behind a VPLEX.

TIA