Bug 840746
| Summary: | BETA 2 - expose NetworkOperationsConnectivityTimeout in rhevm-config | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Virtualization Manager | Reporter: | lpeer <lpeer> |
| Component: | ovirt-engine | Assignee: | Moti Asayag <masayag> |
| Status: | CLOSED CURRENTRELEASE | QA Contact: | Meni Yakove <myakove> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 3.1.0 | CC: | dyasny, ecohen, hateya, iheim, lpeer, mavital, mpavlik, Rhev-m-bugs, sgrinber, yeylon, ykaul |
| Target Milestone: | --- | Keywords: | Improvement |
| Target Release: | 3.1.0 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | network | ||
| Fixed In Version: | si14 | Doc Type: | Enhancement |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2012-12-04 20:04:54 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | Network | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
lpeer
2012-07-17 07:07:11 UTC
*** Bug 820879 has been marked as a duplicate of this bug. *** I recall something about complaints that it was too long. But for setup networks it is critical since we do it when the host is not in maintenance. What will happen if this time is longer then the response timeout? Will the host be fenced? In any case I don't think it should be in the dialogue, but as a rhevm-config parameter. I recall that VDSM did the roll back, so does VSDM supports that on the API? (In reply to comment #4) > I recall something about complaints that it was too long. > > But for setup networks it is critical since we do it when the host is not in > maintenance. What will happen if this time is longer then the response > timeout? Will the host be fenced? In any case I don't think it should be in > the dialogue, but as a rhevm-config parameter. after 3 minutes of not having connectivity to the host, we get a timeout and the host is moved to non-responsive, the regular non-responsive treatment is used. So we don't need to set timeout in the dialog? We already have it in config but this is not editable by the user. We can change this to be available for the users instead of having it in the dialog. simon please approve. > after 3 minutes of not having connectivity to the host, we get a timeout and > the host is moved to non-responsive, the regular non-responsive treatment is > used. That is not good enough, we need VDSM roll back, not a fencing treatment. Is there a VDSM rollback on loss of connectivity? Logic should not change in the engine, if VDSM does fail to roll back then indeed engine should go for non-responsive treatment. Note that this means that the rollback timeout must be less then the fencing trigger - you don't want to get into races here. > > So we don't need to set timeout in the dialog? > > We already have it in config but this is not editable by the user. > We can change this to be available for the users instead of having it in the > dialog. Yes, let's make it configurable by the user, I don't see it necessary to have this parameter editable per setupNetworks operation. > After a discussion with Simon we agreed to add this as a config parameter available for the user but not on the UI dialogs. A suggested patch: http://gerrit.ovirt.org/7006 Tested on rhevm-3.1.0-12.el6ev.noarch, vdsm-4.9.6-28.0.el6_3.x86_64. |