Re: [ovirt-users] Ovirt 4.2 and host console
Il 22 Dic 2017 10:12 PM, "Yaniv Kaul" <ykaul@redhat.com> ha scritto: On Dec 22, 2017 7:33 PM, "Gianluca Cecchi" <gianluca.cecchi@gmail.com> wrote: Hello, after upgrading engine and then plain CentOS 7.4 host from 4.1 to 4.2, I see in host section if I select line for the host, right click and host console... That tries to go to the typical 9090 cockpit Port of node-ng... Is this an error or in 4.2 the access to host console is for plain OS nodes too? In that case is there any service I have to enable on host? It seems indeed my host is not currently listening on 9090 Port.... Cockpit + firewall settings to enable to get to it. Cockpit service should be up and running after the upgrade. Ovirt hist depliy takes care of it. Firewall is configured by the engine unless you disabled firewall config on the host configuration dialog. Didi, can you help here? Gianluca, can you share host upgrade logs? Y. Thanks, Gianluca _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
On Fri, Dec 22, 2017 at 10:22 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il 22 Dic 2017 10:12 PM, "Yaniv Kaul" <ykaul@redhat.com> ha scritto:
On Dec 22, 2017 7:33 PM, "Gianluca Cecchi" <gianluca.cecchi@gmail.com> wrote:
Hello, after upgrading engine and then plain CentOS 7.4 host from 4.1 to 4.2, I see in host section if I select line for the host, right click and host console... That tries to go to the typical 9090 cockpit Port of node-ng... Is this an error or in 4.2 the access to host console is for plain OS nodes too? In that case is there any service I have to enable on host? It seems indeed my host is not currently listening on 9090 Port....
Cockpit + firewall settings to enable to get to it.
Cockpit service should be up and running after the upgrade. Ovirt hist depliy takes care of it. Firewall is configured by the engine unless you disabled firewall config on the host configuration dialog.
Didi, can you help here? Gianluca, can you share host upgrade logs?
Hello, this is a plain CentOS, not ovirt-node-ng one, that was upgraded from 4.1.7 to 4.2 So the upgrade path has been to put the host into maintenance, yum update, reboot. Indeed cockpit has been installed as part of the yum update part: Dec 22 10:40:07 Installed: cockpit-bridge-155-1.el7.centos.x86_64 Dec 22 10:40:07 Installed: cockpit-system-155-1.el7.centos.noarch Dec 22 10:40:40 Installed: cockpit-networkmanager-155-1.el7.centos.noarch Dec 22 10:40:42 Installed: cockpit-ws-155-1.el7.centos.x86_64 Dec 22 10:40:43 Installed: cockpit-155-1.el7.centos.x86_64 Dec 22 10:40:52 Installed: cockpit-storaged-155-1.el7.centos.noarch Dec 22 10:40:57 Installed: cockpit-dashboard-155-1.el7.centos.x86_64 Dec 22 10:41:35 Installed: cockpit-ovirt-dashboard-0.11.3-0.1.el7.centos.noarch I also see that there is a systemd cockpit.service unit that is configured as static and requires a cockpit.socket unit, that in turn is WantedBy sockets.target But if I run [root@ovirt01 ~]# remotectl certificate remotectl: No certificate found in dir: /etc/cockpit/ws-certs.d [root@ovirt01 ~]# So it seems that the cockpit.service ExecStartPre has not been ever run... ExecStartPre=/usr/sbin/remotectl certificate --ensure --user=root --group=cockpit-ws --selinux-type=etc_t [root@ovirt01 ~]# systemctl status cockpit.service ● cockpit.service - Cockpit Web Service Loaded: loaded (/usr/lib/systemd/system/cockpit.service; static; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) [root@ovirt01 ~]# Gianluca
On Sat, Dec 23, 2017 at 11:18 AM, Gianluca Cecchi <gianluca.cecchi@gmail.com
wrote:
On Fri, Dec 22, 2017 at 10:22 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il 22 Dic 2017 10:12 PM, "Yaniv Kaul" <ykaul@redhat.com> ha scritto:
On Dec 22, 2017 7:33 PM, "Gianluca Cecchi" <gianluca.cecchi@gmail.com> wrote:
Hello, after upgrading engine and then plain CentOS 7.4 host from 4.1 to 4.2, I see in host section if I select line for the host, right click and host console... That tries to go to the typical 9090 cockpit Port of node-ng... Is this an error or in 4.2 the access to host console is for plain OS nodes too? In that case is there any service I have to enable on host? It seems indeed my host is not currently listening on 9090 Port....
Cockpit + firewall settings to enable to get to it.
Cockpit service should be up and running after the upgrade. Ovirt hist depliy takes care of it. Firewall is configured by the engine unless you disabled firewall config on the host configuration dialog.
Didi, can you help here? Gianluca, can you share host upgrade logs?
Hello, this is a plain CentOS, not ovirt-node-ng one, that was upgraded from 4.1.7 to 4.2 So the upgrade path has been to put the host into maintenance, yum update, reboot.
Indeed cockpit has been installed as part of the yum update part:
Dec 22 10:40:07 Installed: cockpit-bridge-155-1.el7.centos.x86_64 Dec 22 10:40:07 Installed: cockpit-system-155-1.el7.centos.noarch Dec 22 10:40:40 Installed: cockpit-networkmanager-155-1.el7.centos.noarch Dec 22 10:40:42 Installed: cockpit-ws-155-1.el7.centos.x86_64 Dec 22 10:40:43 Installed: cockpit-155-1.el7.centos.x86_64 Dec 22 10:40:52 Installed: cockpit-storaged-155-1.el7.centos.noarch Dec 22 10:40:57 Installed: cockpit-dashboard-155-1.el7.centos.x86_64 Dec 22 10:41:35 Installed: cockpit-ovirt-dashboard-0.11. 3-0.1.el7.centos.noarch
I also see that there is a systemd cockpit.service unit that is configured as static and requires a cockpit.socket unit, that in turn is WantedBy sockets.target
But if I run
[root@ovirt01 ~]# remotectl certificate remotectl: No certificate found in dir: /etc/cockpit/ws-certs.d [root@ovirt01 ~]#
So it seems that the cockpit.service ExecStartPre has not been ever run... ExecStartPre=/usr/sbin/remotectl certificate --ensure --user=root --group=cockpit-ws --selinux-type=etc_t
[root@ovirt01 ~]# systemctl status cockpit.service ● cockpit.service - Cockpit Web Service Loaded: loaded (/usr/lib/systemd/system/cockpit.service; static; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) [root@ovirt01 ~]#
Gianluca
I see these lines in /var/log/messages (on host I have no iptables and no firewalld active and selinux is permissive): Dec 22 10:40:40 ovirt01 yum[116794]: Installed: cockpit-networkmanager-155-1.el7.centos.noarch Dec 22 10:40:41 ovirt01 systemd: Reloading. Dec 22 10:40:41 ovirt01 vdsm-tool: module dump_volume_chains could not load to vdsm-tool: Traceback (most recent call last):#012 File "/usr/bin/vdsm-tool", line 91, in load_modules#012 mod_absp, mod_desc)#012 File "/usr/lib/python2.7/site-packages/vdsm/tool/dump_volume_chains.py", line 26, in <module>#012 from vdsm import client#012 File "/usr/lib/python2.7/site-packages/vdsm/client.py", line 106, in <module>#012 from vdsm.api import vdsmapi#012 File "/usr/lib/python2.7/site-packages/vdsm/api/vdsmapi.py", line 29, in <module>#012 from vdsm.common.logutils import Suppressed#012ImportError: No module named logutils Dec 22 10:40:41 ovirt01 systemd: [/usr/lib/systemd/system/ip6tables.service:3] Failed to add dependency on syslog.target,iptables.service, ignoring: Invalid argument Dec 22 10:40:41 ovirt01 systemd: Binding to IPv6 address not available since kernel does not support IPv6. ... Dec 22 10:40:43 ovirt01 yum[116794]: Installed: cockpit-155-1.el7.centos.x86_64 Dec 22 10:40:43 ovirt01 yum[116794]: Updated: libmount-2.23.2-43.el7_4.2.x86_64 Dec 22 10:40:43 ovirt01 vdsmd_init_common.sh: vdsm: Running upgraded_version_check Dec 22 10:40:43 ovirt01 vdsmd_init_common.sh: vdsm: Running check_is_configured Dec 22 10:40:43 ovirt01 vdsm-tool: module dump_volume_chains could not load to vdsm-tool: Traceback (most recent call last):#012 File "/usr/bin/vdsm-tool", line 91, in load_modules#012 mod_absp, mod_desc)#012 File "/usr/lib/python2.7/site-packages/vdsm/tool/dump_volume_chains.py", line 26, in <module>#012 from vdsm import client#012 File "/usr/lib/python2.7/site-packages/vdsm/client.py", line 106, in <module>#012 from vdsm.api import vdsmapi#012 File "/usr/lib/python2.7/site-packages/vdsm/api/vdsmapi.py", line 29, in <module>#012 from vdsm.common.logutils import Suppressed#012ImportError: No module named logutils ... Dec 22 10:40:57 ovirt01 yum[116794]: Installed: cockpit-dashboard-155-1.el7.centos.x86_64 Dec 22 10:40:57 ovirt01 vdsm-tool: module dump_volume_chains could not load to vdsm-tool: Traceback (most recent call last):#012 File "/usr/bin/vdsm-tool", line 91, in load_modules#012 mod_absp, mod_desc)#012 File "/usr/lib/python2.7/site-packages/vdsm/tool/dump_volume_chains.py", line 26, in <module>#012 from vdsm import client#012 File "/usr/lib/python2.7/site-packages/vdsm/client.py", line 106, in <module>#012 from vdsm.api import vdsmapi#012 File "/usr/lib/python2.7/site-packages/vdsm/api/vdsmapi.py", line 29, in <module>#012 from vdsm.common.logutils import Suppressed#012ImportError: No module named logutils
On Sat, Dec 23, 2017 at 11:18 AM, Gianluca Cecchi <gianluca.cecchi@gmail.com
wrote:
On Fri, Dec 22, 2017 at 10:22 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il 22 Dic 2017 10:12 PM, "Yaniv Kaul" <ykaul@redhat.com> ha scritto:
On Dec 22, 2017 7:33 PM, "Gianluca Cecchi" <gianluca.cecchi@gmail.com> wrote:
Hello, after upgrading engine and then plain CentOS 7.4 host from 4.1 to 4.2, I see in host section if I select line for the host, right click and host console... That tries to go to the typical 9090 cockpit Port of node-ng... Is this an error or in 4.2 the access to host console is for plain OS nodes too? In that case is there any service I have to enable on host? It seems indeed my host is not currently listening on 9090 Port....
Cockpit + firewall settings to enable to get to it.
Cockpit service should be up and running after the upgrade. Ovirt hist depliy takes care of it. Firewall is configured by the engine unless you disabled firewall config on the host configuration dialog.
Didi, can you help here? Gianluca, can you share host upgrade logs?
Hello, this is a plain CentOS, not ovirt-node-ng one, that was upgraded from 4.1.7 to 4.2 So the upgrade path has been to put the host into maintenance, yum update, reboot.
Indeed cockpit has been installed as part of the yum update part:
Dec 22 10:40:07 Installed: cockpit-bridge-155-1.el7.centos.x86_64 Dec 22 10:40:07 Installed: cockpit-system-155-1.el7.centos.noarch Dec 22 10:40:40 Installed: cockpit-networkmanager-155-1.el7.centos.noarch Dec 22 10:40:42 Installed: cockpit-ws-155-1.el7.centos.x86_64 Dec 22 10:40:43 Installed: cockpit-155-1.el7.centos.x86_64 Dec 22 10:40:52 Installed: cockpit-storaged-155-1.el7.centos.noarch Dec 22 10:40:57 Installed: cockpit-dashboard-155-1.el7.centos.x86_64 Dec 22 10:41:35 Installed: cockpit-ovirt-dashboard-0.11. 3-0.1.el7.centos.noarch
I also see that there is a systemd cockpit.service unit that is configured as static and requires a cockpit.socket unit, that in turn is WantedBy sockets.target
But if I run
[root@ovirt01 ~]# remotectl certificate remotectl: No certificate found in dir: /etc/cockpit/ws-certs.d [root@ovirt01 ~]#
So it seems that the cockpit.service ExecStartPre has not been ever run... ExecStartPre=/usr/sbin/remotectl certificate --ensure --user=root --group=cockpit-ws --selinux-type=etc_t
[root@ovirt01 ~]# systemctl status cockpit.service ● cockpit.service - Cockpit Web Service Loaded: loaded (/usr/lib/systemd/system/cockpit.service; static; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) [root@ovirt01 ~]#
Gianluca
At the end I compared my CentOS 7.4 oVirt 4.2 configuration with the configuration of a RHEV 4.1 RHV-H environment and there I see this kind of config [root@rhevora1 ~]# systemctl status cockpit.socket ● cockpit.socket - Cockpit Web Service Socket Loaded: loaded (/usr/lib/systemd/system/cockpit.socket; enabled; vendor preset: enabled) Active: active (listening) since Fri 2017-10-27 17:02:30 CEST; 1 months 29 days ago Docs: man:cockpit-ws(8) Listen: [::]:9090 (Stream) Oct 27 17:02:30 rhevora1.mydomain systemd[1]: Listening on Cockpit Web Service Socket. Oct 27 17:02:30 rhevora1.mydomain systemd[1]: Starting Cockpit Web Service Socket. [root@rhevora1 ~]# So there is basically the difference that on RHEV RHV-H the cockpit.socket systemd unit is enabled by default (cockpit-ws-148-1.el7.x86_64), while on this just updated to 4.2 oVirt host on plain CentOS we have [root@ovirt01 etc]# systemctl status cockpit.socket ● cockpit.socket - Cockpit Web Service Socket Loaded: loaded (/usr/lib/systemd/system/cockpit.socket; disabled; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) Listen: 0.0.0.0:9090 (Stream) [root@ovirt01 etc]# On oVirt 4.2 the version is currently cockpit-ws-155-1.el7.centos.x86_64 What I've done has been systemctl enable cockpit.socket systemctl start cockpit.socket and now I have the ovirt01 host correctly listening on 9090 port and I'm able to web connect and enjoy... Let me know if this is a sort of regression of cockpit-ws package or what... Gianluca
Hi As you can see from the systemctl status, the difference is caused due to the preset rule for cockpit.socket, which is enabled on the node-ng based system, and disabled on your regular centos based system. If you take a look at the cockpit postinstall script, you'd see something like: systemctl --no-reload preset cockpit.socket... This sets the cockpit.socket according to its preset rule. Thanks, Yuval. On Tue, Dec 26, 2017 at 3:35 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
On Sat, Dec 23, 2017 at 11:18 AM, Gianluca Cecchi < gianluca.cecchi@gmail.com> wrote:
On Fri, Dec 22, 2017 at 10:22 PM, Sandro Bonazzola <sbonazzo@redhat.com> wrote:
Il 22 Dic 2017 10:12 PM, "Yaniv Kaul" <ykaul@redhat.com> ha scritto:
On Dec 22, 2017 7:33 PM, "Gianluca Cecchi" <gianluca.cecchi@gmail.com> wrote:
Hello, after upgrading engine and then plain CentOS 7.4 host from 4.1 to 4.2, I see in host section if I select line for the host, right click and host console... That tries to go to the typical 9090 cockpit Port of node-ng... Is this an error or in 4.2 the access to host console is for plain OS nodes too? In that case is there any service I have to enable on host? It seems indeed my host is not currently listening on 9090 Port....
Cockpit + firewall settings to enable to get to it.
Cockpit service should be up and running after the upgrade. Ovirt hist depliy takes care of it. Firewall is configured by the engine unless you disabled firewall config on the host configuration dialog.
Didi, can you help here? Gianluca, can you share host upgrade logs?
Hello, this is a plain CentOS, not ovirt-node-ng one, that was upgraded from 4.1.7 to 4.2 So the upgrade path has been to put the host into maintenance, yum update, reboot.
Indeed cockpit has been installed as part of the yum update part:
Dec 22 10:40:07 Installed: cockpit-bridge-155-1.el7.centos.x86_64 Dec 22 10:40:07 Installed: cockpit-system-155-1.el7.centos.noarch Dec 22 10:40:40 Installed: cockpit-networkmanager-155-1.el7.centos.noarch Dec 22 10:40:42 Installed: cockpit-ws-155-1.el7.centos.x86_64 Dec 22 10:40:43 Installed: cockpit-155-1.el7.centos.x86_64 Dec 22 10:40:52 Installed: cockpit-storaged-155-1.el7.centos.noarch Dec 22 10:40:57 Installed: cockpit-dashboard-155-1.el7.centos.x86_64 Dec 22 10:41:35 Installed: cockpit-ovirt-dashboard-0.11.3 -0.1.el7.centos.noarch
I also see that there is a systemd cockpit.service unit that is configured as static and requires a cockpit.socket unit, that in turn is WantedBy sockets.target
But if I run
[root@ovirt01 ~]# remotectl certificate remotectl: No certificate found in dir: /etc/cockpit/ws-certs.d [root@ovirt01 ~]#
So it seems that the cockpit.service ExecStartPre has not been ever run... ExecStartPre=/usr/sbin/remotectl certificate --ensure --user=root --group=cockpit-ws --selinux-type=etc_t
[root@ovirt01 ~]# systemctl status cockpit.service ● cockpit.service - Cockpit Web Service Loaded: loaded (/usr/lib/systemd/system/cockpit.service; static; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) [root@ovirt01 ~]#
Gianluca
At the end I compared my CentOS 7.4 oVirt 4.2 configuration with the configuration of a RHEV 4.1 RHV-H environment and there I see this kind of config
[root@rhevora1 ~]# systemctl status cockpit.socket ● cockpit.socket - Cockpit Web Service Socket Loaded: loaded (/usr/lib/systemd/system/cockpit.socket; enabled; vendor preset: enabled) Active: active (listening) since Fri 2017-10-27 17:02:30 CEST; 1 months 29 days ago Docs: man:cockpit-ws(8) Listen: [::]:9090 (Stream)
Oct 27 17:02:30 rhevora1.mydomain systemd[1]: Listening on Cockpit Web Service Socket. Oct 27 17:02:30 rhevora1.mydomain systemd[1]: Starting Cockpit Web Service Socket. [root@rhevora1 ~]#
So there is basically the difference that on RHEV RHV-H the cockpit.socket systemd unit is enabled by default (cockpit-ws-148-1.el7.x86_64), while on this just updated to 4.2 oVirt host on plain CentOS we have
[root@ovirt01 etc]# systemctl status cockpit.socket ● cockpit.socket - Cockpit Web Service Socket Loaded: loaded (/usr/lib/systemd/system/cockpit.socket; disabled; vendor preset: disabled) Active: inactive (dead) Docs: man:cockpit-ws(8) Listen: 0.0.0.0:9090 (Stream) [root@ovirt01 etc]#
On oVirt 4.2 the version is currently cockpit-ws-155-1.el7.centos.x86_64
What I've done has been systemctl enable cockpit.socket systemctl start cockpit.socket
and now I have the ovirt01 host correctly listening on 9090 port and I'm able to web connect and enjoy...
Let me know if this is a sort of regression of cockpit-ws package or what...
Gianluca
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
participants (3)
-
Gianluca Cecchi -
Sandro Bonazzola -
Yuval Turgeman