High latency on VNX5300 with multipath and ALUA

All hosts are connected to the same array, as follow:

host A (multipath) ----> FC switch ----> VNX5300 ----> ERROR

host B (multipath) ----> FCOE switch ----> VNX5300 --> ERROR

OS: centos6 and centos7

All hosts are up to date and this is our multipath configuration suggested by EMC:

# RHEV REVISION 1.1

# RHEV PRIVATE

devices {

device {

    vendor                  "DGC"

    product                 ".*"

    product_blacklist       "LUNZ"

    path_grouping_policy    group_by_prio

    path_selector           "round-robin 0"

    path_checker            emc_clariion

    features                "1 queue_if_no_path"

    hardware_handler        "1 alua"

    prio                    alua

    failback                immediate

    rr_weight               uniform

    no_path_retry           60

    rr_min_io               1

}

}

The command multipath -ll hung when executed on the host while the problem occur (nothing logged in /var/log/messages or dmesg).

I tested latency with ioping:

ioping /dev/6a386652-629d-4045-835b-21d2f5c104aa/metadata

Usually it return "time=15.6 ms", sometimes return "time=19 s" (yes, seconds)

Any idea or suggestion?

Thanks

Stefano