Description of problem: Deploying the first hosted-engine host, after the management bridge creation, the default gateway is not restored. As consequence, in this case, the deploy fails due to the missing route to the NFS storage, which is in a different network. This issue was observed when using a management bridge different from "rhevm", trough the option "OVEHOSTED_NETWORK/bridgeName=str:br0" in answer file. Pre-deploy: [root@host1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-eth0 DEVICE="eth0" BOOTPROTO="static" DNS1=192.168.122.1 GATEWAY=192.168.122.1 IPADDR=192.168.122.207 IPV6INIT="no" MTU="1500" NETMASK="255.255.255.0" NM_CONTROLLED="no" ONBOOT="yes" TYPE="Ethernet" [root@host1 ~]# route -n Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 192.168.122.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 0.0.0.0 192.168.122.1 0.0.0.0 UG 0 0 0 eth0 [root@host1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-br0 cat: /etc/sysconfig/network-scripts/ifcfg-br0: No such file or directory [root@host1 ~]# brctl show bridge name bridge id STP enabled interfaces [root@host1 ~]# Deploy: [root@host1 ~]# hosted-engine --deploy --config-append=/tmp/answers.conf ... --== CONFIGURATION PREVIEW ==-- Bridge interface : eth0 Engine FQDN : hostedengine.pahim.org Bridge name : br0 SSH daemon port : 22 Firewall manager : iptables Gateway address : 192.168.122.1 Host name for web application : host1.pahim.org Host ID : 1 Image alias : hosted_engine Image size GB : 25 Storage connection : 192.168.25.118:/nfs/hostedengine Console type : vnc Memory size MB : 2048 MAC address : 00:16:3e:78:23:42 Boot type : cdrom Number of CPUs : 2 ISO image (for cdrom boot) : /tmp/RHEL6.4-20130130.0-Server-x86_64-DVD1.iso CPU Type : model_SandyBridge [ INFO ] Stage: Transaction setup [ INFO ] Stage: Misc configuration [ INFO ] Stage: Package installation [ INFO ] Stage: Misc configuration [ INFO ] Configuring libvirt [ INFO ] Configuring VDSM [ INFO ] Starting vdsmd [ INFO ] Waiting for VDSM hardware info [ INFO ] Configuring the management bridge [ ERROR ] Failed to execute stage 'Misc configuration': Connection to storage server failed [ INFO ] Stage: Clean up [ INFO ] Generating answer file '/var/lib/ovirt-hosted-engine-setup/answers/answers-20150525135306.conf' [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination ... vdsm.log: Thread-15::DEBUG::2015-05-25 13:53:00,567::BindingXMLRPC::1133::vds::(wrapper) client [127.0.0.1]::call setupNetworks with ({'br0': {'nic': 'eth0', 'netmask': '255.255.255.0', 'bootproto': 'static', 'ipaddr': '192.168.122.207', 'gateway': '192.168.122.1'}}, {}, {'connectivityCheck': False}) {} Thread-15::DEBUG::2015-05-25 13:53:05,435::BindingXMLRPC::1140::vds::(wrapper) return setupNetworks with {'status': {'message': 'Done', 'code': 0}} [root@host1 ~]# ip a 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 52:54:00:bd:cd:e7 brd ff:ff:ff:ff:ff:ff 5: ;vdsmdummy;: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN link/ether ee:33:dc:74:04:52 brd ff:ff:ff:ff:ff:ff 6: bond0: <BROADCAST,MULTICAST,MASTER> mtu 1500 qdisc noop state DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 8: br0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN link/ether 52:54:00:bd:cd:e7 brd ff:ff:ff:ff:ff:ff inet 192.168.122.207/24 brd 192.168.122.255 scope global br0 inet6 fe80::5054:ff:febd:cde7/64 scope link valid_lft forever preferred_lft forever [root@host1 ~]# route -n Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 192.168.122.0 0.0.0.0 255.255.255.0 U 0 0 0 br0 169.254.0.0 0.0.0.0 255.255.0.0 U 1008 0 0 br0 [root@host1 ~]# cat /var/lib/vdsm/persistence/netconf/nets/br0 {"nic": "eth0", "netmask": "255.255.255.0", "bootproto": "static", "ipaddr": "192.168.122.207", "gateway": "192.168.122.1"} [root@host1 ~]# cat /var/run/vdsm/netconf/nets/br0 {"nic": "eth0", "netmask": "255.255.255.0", "bootproto": "static", "ipaddr": "192.168.122.207", "gateway": "192.168.122.1"} [root@host1 ~]# brctl show bridge name bridge id STP enabled interfaces ;vdsmdummy; 8000.000000000000 no br0 8000.525400bdcde7 no eth0 Version-Release number of selected component (if applicable): ovirt-hosted-engine-setup-1.2.2-3.el6ev.noarch ovirt-hosted-engine-ha-1.2.5-1.el6ev.noarch vdsm-4.16.13.1-1.el6ev.x86_64 Additional info: Attaching setup log file.
Created attachment 1029545 [details] ovirt-hosted-engine-setup.log
Adding a custom named management network is not supported in 3.5. It will be added in 3.6.0. Moving to VERIFY on 3.6.
Current status on 3.6 rc1 upstream: The user can run hosted-engine-setup passing an answer-file with a custom name for the management bridge. hosted-engine-setup will create the bridge with that name. The default gateway is correctly set regardless of the management bridge name. oVirt engine on the engine VM is by default trying to use a logical network called ovirtmgmt so it's then trying to bind it to a bridge called that way. Cause it doesn't found it, the host will fail to come up. Then hosted-engine-setup discovers that the host is in a non-operation state so it advises the user to manually fix it into the engine before retrying. [ INFO ] Waiting for the host to become operational in the engine. This may take several minutes... [ INFO ] Still waiting for VDSM host to become operational... The host hosted_engine_1 is in non-operational state. Please try to activate it via the engine webadmin UI. Retry checking host status or ignore this and continue (Retry, Ignore)[Retry]? retry ovirt-hosted-engine-setup doesn't provide any helpful hint to the user on this point. If the user manually: - connects to the engine webui - renames the management network for the selected datacenter - manually activates the host the host comes up in the engine and ovirt-hosted-engine-setup could continue.
Danken, if we know on hosted-engine side that the user want to use a custom name for its management network on a specific cluster, is there any API to pragmatically tell the engine to rename that management network to the desired value?
Can you please check the status of this ticket? The customer ticket doesn't seem to be related to 3.6.
Removing the customer case since it's unrelated to 3.6
Setting management network is possible on cluster level during cluster creating or later on in case the cluster is empty (no hosts attached to it) by updating the cluster. When creating/updating a cluster through REST API "management_network" element could be supplied in order to specify the desired management network. Please note that the desired management network should be created in the DC prior creating/updating the cluster. Here are the relevant URL's: Create network: POST ovirt-engine/api/datacenters/{datacenter:id}/networks Create cluster: POST ovirt-engine/api/datacenters/{datacenter:id}/clusters Update cluster: PUT ovirt-engine/api/datacenters/{datacenter:id}/clusters/{cluster:id} For the creating network URL "network" element should be supplied and for cluster related URL's "cluster" element. You can obtain schema that defines the XML structure that should be supplied with the URL's by accessing ovirt-engine/api?schema and RSDL is available by ovirt-engine/api?rsdl URL.
*** This bug has been marked as a duplicate of bug 1231799 ***