Bug 822930
Summary: | DNS info is removed after register to rhevm | ||||||
---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 6 | Reporter: | shine <shcao> | ||||
Component: | ovirt-node | Assignee: | Fabian Deutsch <fdeutsch> | ||||
Status: | CLOSED ERRATA | QA Contact: | Virtualization Bugs <virt-bugs> | ||||
Severity: | urgent | Docs Contact: | |||||
Priority: | urgent | ||||||
Version: | 6.3 | CC: | acathrow, bsarathy, cshao, dougsland, fdeutsch, gouyang, jboggs, leiwang, mburns, moli, ovirt-maint, sgordon, syeghiay, yuachen | ||||
Target Milestone: | rc | ||||||
Target Release: | --- | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Whiteboard: | |||||||
Fixed In Version: | ovirt-node-2.3.0-8.el6 | Doc Type: | Bug Fix | ||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | |||||||
: | 824621 (view as bug list) | Environment: | |||||
Last Closed: | 2012-07-19 14:13:06 UTC | Type: | Bug | ||||
Regression: | --- | Mount Type: | --- | ||||
Documentation: | --- | CRM: | |||||
Verified Versions: | Category: | --- | |||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
Cloudforms Team: | --- | Target Upstream Version: | |||||
Embargoed: | |||||||
Bug Depends On: | |||||||
Bug Blocks: | 824621 | ||||||
Attachments: |
|
the root cause is that dns is removed after register to rhevm, to reproduce this can specify the dns on tui manually and then register to rhevm. set priority to urgent to highlight this issue. Mburns, Please help assign the bug to vdsm component it it's not owned by ovirt-node. I looking into this some more and it's not exactly vdsm causing the problem (though that mght be where the fix is). Sequence: configure network set different dns apply changes drop to shell check /etc/resolv.conf -- setup correctly service network restart check /etc/resolv.conf -- empty The rhevm registration does a service network restart which might be where we need to fix this Mike, I could not reproduce the problem by following your steps. In my tests /etc/resolv.conf was never empty. I configured the NIC with DNS and used the IP of a non-existent DNS server. Okay, I finally could reproduce this. The problem is the following: The network is configured using DHCP, dhclient creates a copy of the original (a) /etc/resolv.conf in e.g. (b) /var/lib/dhclient/resolv.conf.predhclient.breth0 . When dhclient is shut down, it overwrites /etc/resolv.conf with (b), thus, any changes that happened during the lifetime of dhclient to /etc/resolv.conf are lost. If /etc/resolv.conf was empty before dhclient was started, and a DNS server is manually given in the TUI while dhclient is running, then - when a network restart happens - /etc/resolv.conf will be empty again and we've got this problem. AFAIU the same applies to the ntp configuration. The following patch should fix this problem: http://gerrit.ovirt.org/4674 Verified with rhevh-6.3-20120523.1. the DNS won't be removed after restarting network service. It also successes to register to rhev-m and do vm migration. The tech note for this was covered in a related bug, so no technical note required for this one. Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2012-0741.html |
Created attachment 585425 [details] VMMigError Description of problem: VM migration failed with empty resolv.conf after registered to RHEVM Version-Release number of selected component: rhev-hypervisor6-6.3-20120516.0.el6 How reproducible: 100% Not Regression Priority: High Steps to Reproduce: 1. install rhev-hypervisor6-6.3-20120516.0.el6 to the machine 2. register it to rhevm 3. find the /etc/resolv.conf was empty 4. Connect to NFS storage, and create VMs on the RHEVH host Actual results: resolv.conf is empty, which results in DNS error and vm migration failure.[migerrorvdsm.log] But the resolv.conf wouldn't be empty, after rebooted the rhevh. Then vm migration still failed. If the hostnames of srchost and desthost are different, I modified the host-name setted on rhevm as the hostname of RHEVH, the migration is successful.[pic successfulmig.png] Expected results: vm migration should be successful. Additional info: