Bug 746745

Summary: non descriptive error message when trying to use already used IP
Product: [Retired] oVirt Reporter: Michael Brown <michael>
Component: ovirt-engine-coreAssignee: lpeer <lpeer>
Status: CLOSED WONTFIX QA Contact:
Severity: high Docs Contact:
Priority: medium    
Version: unspecifiedCC: acathrow, danken, dfediuck, ecohen, iheim, knesenko, leiwang, lpeer, mburns, ovirt-maint, Rhev-m-bugs, ycui, yeylon, ykaul
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard: network
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-02-25 07:24:54 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Bug Depends On: 739744    
Bug Blocks:    
Attachments:
Description Flags
Screenshot of actual error
none
rhevm.log and vdsm.log as requested none

Description Michael Brown 2011-10-17 16:43:54 UTC 
Created attachment 528599 [details]
Screenshot of actual error

Description of problem:
RHEV IP address validation rules are incorrect


Version-Release number of selected component (if applicable):
RHEV Manager for Servers and Desktops: 3.0.0_0001-41.el6

Red Hat Enterprise Virtualization Hypervisor Hosts:
        rhev1.netdirect.ca:     OS Version - RHEL - 6Server - 6.1.0.2.el6       VDSM Version - 3.0.96.1
        rhev2.netdirect.ca:     OS Version - RHEL - 6Server - 6.1.0.2.el6       VDSM Version - 3.0.96.1
        rhev3.netdirect.ca:     OS Version - RHEL - 6Server - 6.1.0.2.el6       VDSM Version - 3.0.96.1


How reproducible:
Consistently failed, needed to set a different IP and change manually

Steps to Reproduce:
1. Attempt to set IP address containing a zero
2. Watch it fail (see screenshot)
  
Actual results:
IP address validation rejects good IP addresses

Expected results:
IP address should be accepted as valid and configured on the hypervisor.

Additional info:
Other addresses that should be accepted but probably aren't:
192.168.1.0/23
192.168.0.255/23

See the explanation and guide here:
http://serverfault.com/questions/290889/is-it-valid-to-have-part-of-an-ipv4-address-set-to-zero/290891#290891
Comment 2 Einav Cohen 2011-10-17 21:34:29 UTC 
According to the screen-shot, there is no GUI validation failure or even a CanDoAction backend failure - the failure occurs in the ExecuteAction phase.

Not sure if Backend or VDSM are responsible for the failure -> moving the BZ to the Backend component for further investigation.

Michael - I suggest attaching here the rhevm.log and the vdsm.log.
Comment 3 Michael Brown 2011-10-18 01:43:48 UTC 
Created attachment 528679 [details]
rhevm.log and vdsm.log as requested
Comment 4 Kiril Nesenko 2011-10-18 08:02:42 UTC 
Looks like you trying to use ip address that is already in use, please check the logs and your network configuration:

2011-10-13 20:38:33,752 INFO  [org.nogah.vdsbroker.vdsbroker.AddNetworkVDSCommand] (http-0.0.0.0-8443-1) START, AddNetworkVDSCommand(vdsId = 1801621c-f5f4-11e0-9422-00215e978220, networkName=rhevm, oldNetworkName=null, hostAddr=null, checkConnectivity=false, connectionTimeout=0, vlanId=null, bondName=null, nics=[eth0], inetAddr=192.168.0.62, networkMask=255.255.255.0, gateway=, stp=false, bondingOptions=null, bootProtocol=StaticIp), log id: 24b88ae4
2011-10-13 20:38:33,813 INFO  [org.nogah.vdsbroker.vdsbroker.BrokerCommandBase] (http-0.0.0.0-8443-1) Command org.nogah.vdsbroker.vdsbroker.AddNetworkVDSCommand return value 
 Class Name: org.nogah.vdsbroker.vdsbroker.StatusOnlyReturnForXmlRpc
mStatus                       Class Name: org.nogah.vdsbroker.vdsbroker.StatusForXmlRpc
mCode                         22
mMessage                      IP address is already in use
Comment 5 Itamar Heim 2011-10-18 09:20:20 UTC 
even if this is "admin error", we should keep this bug open to fix the fact we don't report the error to the admin and the message only appears in the log
Comment 6 Doron Fediuck 2011-10-18 10:02:37 UTC 
(In reply to comment #5)
> even if this is "admin error", we should keep this bug open to fix the fact we
> don't report the error to the admin and the message only appears in the log

vdsm generic error 22 is being translated to ERR_BAD_ADDR, which has the
text of-ERR_BAD_ADDR=Illegal or Incomplete IP Address.
and this is what the admin gets.
In order to resolve it you need to split vdsm error code to in-use / other
and then give it the relevant text.
So this should be split into a vdsm bz and backend bz, and you need to decide
which version / milestone to target.
Comment 8 Dan Kenigsberg 2011-10-18 11:54:25 UTC 
Vdsm returned a generic error code 22, with a specific textual message "IP address is already in use". It seems that Doron is right and rhev-m simply ignores the text, replacing it with its own localized version of 22's generic text. Having a specific error code for this case is a reasonable request for 3.1.
Comment 9 Michael Brown 2011-10-18 14:30:24 UTC 
(In reply to comment #4)
> Looks like you trying to use ip address that is already in use, please check
> the logs and your network configuration:

In this particular case, that IP address was already on the machine in question and I was reconfiguring interfaces without changing the IP.
Comment 10 Itamar Heim 2013-02-25 07:24:54 UTC 
Closing old bugs. If this issue is still relevant/important in current version, please re-open the bug.