On Wed, Sep 23, 2020 at 1:22 PM Ales Musil <amusil@redhat.com> wrote:

On Wed, Sep 23, 2020 at 1:18 PM Nir Soffer <nsoffer@redhat.com> wrote:
On Wed, Sep 23, 2020 at 1:29 PM Gianluca Cecchi
<gianluca.cecchi@gmail.com> wrote:
>
> Hello,
> supposing to have a node that connects to an iSCSI storage domain in oVirt 4.4, is there any particular requirement in the configuration of the network adapter (ifcfg-eno1 file) when I pre-configure the server OS?
> Eg, do I need to have it managed by NetworkManager in 4.4? Can I instead set NM_CONTROLLED=no for this controller?

4.4 is using NetworkManager, so it is unlikely be able to manage the nic in the
suggested configuration.

Adding Ales to add more info on this.

Nir

As Nir said, 4.4 is using NetworkManager to manage every network on the host.
What would be the reason to disable it? The networking should work the same way at least from
oVirt point of view.

Best regards,
Ales

--

Ales Musil

Software Engineer - RHV Network

Red Hat EMEA

amusil@redhat.com IM: amusil

In this particular case I would like to see if NetworkManager could be in any way responsible for some disconnect/reconnect events we are registering when pushing to the iSCSI storage.

But I'm not sure about it, only a test I would like to do (mainly to exclude it), as I also see NetworkManager registered events about the state of the link.

I'm more inclined to think that the problem could be related to the ixgbe 10gb kernel module shipped with 4.4.1 ovirt node ng.

we see this in messages:

Sep 23 08:00:46 host3 kernel: ixgbe 0000:18:00.1 eno2: NIC Link is Down
Sep 23 08:00:55 host3 kernel: connection1:0: ping timeout of 5 secs expired, recv timeout 5, last rx 4348764427, last ping 4348769472, now 4348774912
Sep 23 08:00:55 host3 kernel: connection1:0: detected conn error (1022)
Sep 23 08:00:55 host3 kernel: sd 15:0:0:0: [sdb] tag#30 FAILED Result: hostbyte=DID_TRANSPORT_DISRUPTED driverbyte=DRIVER_OK cmd_age=6s
Sep 23 08:00:55 host3 kernel: sd 15:0:0:0: [sdb] tag#30 CDB: Test Unit Ready 00 00 00 00 00 00
Sep 23 08:00:55 host3 kernel: sd 15:0:0:1: [sdc] tag#29 FAILED Result: hostbyte=DID_TRANSPORT_DISRUPTED driverbyte=DRIVER_OK cmd_age=6s
Sep 23 08:00:55 host3 kernel: sd 15:0:0:1: [sdc] tag#29 CDB: Test Unit Ready 00 00 00 00 00 00
Sep 23 08:00:55 host3 iscsid[2879]: iscsid: Kernel reported iSCSI connection 1:0 error (1022 - ISCSI_ERR_NOP_TIMEDOUT: A NOP has timed out) state (3)
Sep 23 08:01:00 host3 kernel: session1: session recovery timed out after 5 secs
Sep 23 08:01:00 host3 kernel: sd 15:0:0:0: rejecting I/O to offline device
Sep 23 08:01:00 host3 kernel: blk_update_request: I/O error, dev sdb, sector 1020005680 op 0x1:(WRITE) flags 0xca00 phys_seg 1 prio class 0
Sep 23 08:01:00 host3 kernel: sd 15:0:0:0: rejecting I/O to offline device
Sep 23 08:01:00 host3 kernel: blk_update_request: I/O error, dev sdb, sector 264192 op 0x0:(READ) flags 0x4200 phys_seg 1 prio class 0
Sep 23 08:01:00 host3 kernel: sd 15:0:0:0: rejecting I/O to offline device
Sep 23 08:01:00 host3 kernel: sd 15:0:0:0: rejecting I/O to offline device
Sep 23 08:01:00 host3 multipathd[1919]: sdb: mark as failed
Sep 23 08:01:00 host3 multipathd[1919]: 3600140502d7e875e48740398630ef780: Entering recovery mode: max_retries=4
Sep 23 08:01:00 host3 multipathd[1919]: 3600140502d7e875e48740398630ef780: remaining active paths: 0
Sep 23 08:01:00 host3 kernel: device-mapper: multipath: 253:13: Failing path 8:16.
Sep 23 08:01:00 host3 kernel: sd 15:0:0:1: rejecting I/O to offline device
Sep 23 08:01:00 host3 kernel: blk_update_request: I/O error, dev sdc, sector 264192 op 0x0:(READ) flags 0x4200 phys_seg 1 prio class 0
Sep 23 08:01:00 host3 kernel: sd 15:0:0:1: rejecting I/O to offline device
Sep 23 08:01:00 host3 kernel: device-mapper: multipath: 253:12: Failing path 8:32.
Sep 23 08:01:01 host3 kernel: ixgbe 0000:18:00.1 eno2: NIC Link is Up 10 Gbps, Flow Control: None
Sep 23 08:01:01 host3 NetworkManager[2247]: <info> [1600840861.5116] device (eno2): carrier: link connected

Other 4.4.1 nodes with different network adapters and connected to the same iSCSI storage don't have this problem

Any suggested way to substitute the provided module (5.1.0-k-rh8.2.0) of kernel 4.18.0-193.14.2.el8_2.x86_64 with a more recent one and see if it solves the problems? Or suggestions regarding iscsid.conf customization if using ixgbe based network cards...?

Thanks,

Gianluca