1. Feature Overview: a) Name of feature: Display of NIC/FC fault on RHEV-M management screen b) Feature Description: RHEV-M displays fault information when a fault occurs at the redundant NIC/FC-HBA. As a result we can receive the fault information quickly. 2. Feature Details: a) Architectures: 64-bit Intel EM64T/AMD64 b) Bugzilla Dependencies: None c) Drivers or hardware dependencies: None d) Upstream acceptance information: None e) External links: None f) Severity (H,M,L): High g) Feature Needed by: 2014 1Q
In rhevm-3.3 vdsm reports both states of the bond and its slaves. However, in the 'network interfaces' sub-tab of the host only the bond's slaves status is shown and the user cannot know the bond status. The only method to query the bond state is by rest: /api/hosts/{host:id}/nics/{nic:id}/statistics where {nic:id} should be the bond's id. The bond/slaves state can be reflected to the user by adding: 1. An icon next to the bond name in order to indicate its status. 2. An event log in case on of the bond's slaves is reported 'down' (only for bonds that a network is attached)
Hi Larry, As of 3.3, the bond and slaves link status (UP/DOWN) are reported and reflected in the UI in two places: the host's Interfaces tab and the Setup Host Interfaces dialog. If a slave is going down for some reason, this is reflected in those two places and you can see that the particular NIC is down while the bond is still up. I believe this is cover the original customer request as outlined in comment #1 in this RFE. I am also attaching two scrrenshots form the UI - can you please take a look? As Moti wrote in comment #5, we can add an icon next to the bond name in order to indicate its current status (e.g., UP, UP w/ a slave fault, DOWN), but I think this is not the original customer request.
Switching the NEEDINFO to the repporter.
Administrators do not always monitor the bonding screen (host's Interfaces tab and the Set Up Host Interface dialog from Nir-san's comment12). Then the administrators can not notice a NIC fault at once. Hitachi requests that RHEV-M displays fault messages to the area of Alerts, Events and Task part which is the bottom of the RHEV-M screen because administrators always monitor them. I'll attach RHEV-M screen. Thank you.
(In reply to yuzuru.maya.zn from comment #15) > Administrators do not always monitor the bonding screen > (host's Interfaces tab and the Set Up Host Interface dialog > from Nir-san's comment12). > Then the administrators can not notice a NIC fault at once. > Hitachi requests that RHEV-M displays fault messages > to the area of Alerts, Events and Task part > which is the bottom of the RHEV-M screen > because administrators always monitor them. I'm in favour of this approach as stated in comment #5 (the second proposal). Just to state again, we should report a warning to the event log on for bond which a network is configured on top of them and one or more of their slaves is down.
Can we agree that a warning in the event log is what is missing here? Isn't it what you asked for in Bug 987302?
Let me double check the meaning of "event log". Does the part of event log mean the bottom area of RHEV-M screen? I attach the RHEV-M screen and show "the event log".
(In reply to yuzuru.maya.zn from comment #19) > Let me double check the meaning of "event log". > Does the part of event log mean the bottom area of RHEV-M screen? > I attach the RHEV-M screen and show "the event log". Yes, that's were the events are shown. There is also a specific events view per entity (Host/Vm/...) as well when you click on the specific entity.
*** Bug 987302 has been marked as a duplicate of this bug. ***
Would this satisfy your request: 1) Report a warning to the event log on for bond which a network is configured on top of them and one or more of their slaves is down 2) Add an icon next to the bond name in the Web Admin UI in order to indicate its current status (e.g., UP, UP w/ a slave fault, DOWN) Please note that Ethernet and FC ports are treated differently in the system. Kindly update BZ #1072007 with the FC details so the Storage team would be able to examine the information.
When the status of the interface/bond is changed an event should be reported in the following cases- 1. There is a network or label on the interface/bond 2. The interface is a slave of a bond and 1 is true for the parent bond. There won't be more than one event per interface each 30 minutes. The new events are- HOST_INTERFACE_STATE_UP, HOST_INTERFACE_STATE_DOWN, HOST_BOND_SLAVE_STATE_UP, HOST_BOND_SLAVE_STATE_DOWN The user can subscribe to ovirt-engine-notifier to get notifications on those events via UI or configuration.
verified @ ovirt-3.5.0-rc1
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. https://rhn.redhat.com/errata/RHSA-2015-0158.html