Il giorno Ven 14 Gen 2011 16:07:42 CET, Dejan Muhamedagic ha scritto: [...] > Unfortunately, it cannot really improve situation in any way. > Dropping half the agent and not knowing where is the problem is > not helpful. At least not if you want to share your findings > with the community.
I perfectly know, but my goal for now was to solve my problem, as I repeated (too) many times in these days, I was in trouble with a customer and simplifying the script was the first step to understand how this was made. > One of the possible explanations is that the target simply > wasn't ready at the time it tried to start (as it has only be > made available by the previous service). So, perhaps inserting a > Delay resource in between could help here. If so, then maybe > iSCSI* agents need to make sure that the service is really ready > after start exits. Consider that I have reproduced a virtual environment that can be used for doing all of the tests needed. I don't think that delay is the solution, but I agree with you that maybe reviewing the discovery code issue can help us. > Looking at the diff, though it's simply impossible to figure out > what changed because so many things did, it looks like you > dropped the discovery code (iscsiadm -m discovery). Was that > where you had problems? Well, we'll probably never know for > sure. > Unhappy, > Dejan I can say for sure that we will surely know. As you can see, in all of my posts (and projects) I never give up until there is a clear solution to the problem. I will find out also in this case. Thanks again, -- Raoul Scarazzini Mia Mamma Usa Linux: Niente รจ impossibile da capire, se lo spieghi bene! ra...@miamammausalinux.org _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems