Hi, can nobody shed some light on this? This is reproducible, we always get the exact same error during this command: [org.ovirt.engine.core.utils.ssh.SSHDialog] (pool-6-thread-32) SSH stderr during command root@10.0.1.4:'umask 0077; MYTMP="$(mktemp -t ovirt-XXXXXXXXXX)"; trap "chmod -R u+rwX \"${M YTMP}\" > /dev/null 2>&1; rm -fr \"${MYTMP}\" > /dev/null 2>&1" 0; rm -fr "${MYTMP}" && mkdir "${MYTMP}" && tar --warning=no-timestamp -C "${MYTMP}" -x && "${MYTMP}"/setup DIALOG/dialect=str:machine DIALOG/cust omization=bool:True': stderr: bash: line 1: 8635 Segmentation fault "${MYTMP}"/setup DIALOG/dialect=str:machine DIALOG/customization=bool:True it's an segmentation fault, so it seems to me there must be a bug hiding somewhere. Can anyone confirm to me that what I try to do is supported and/or works? Adding a 2.6.1 oVirt-Node to an 3.3.0-4.el6 Management Node? We are really stuck at this point. This error always shows up when trying to acknowledge the node through the webadmin UI, after registration. Regards Sven On 09.10.2013 11:22, Sven Kieske wrote: > Hi, > > we have successfully deployed ovirt engine 3.3. > However, when adding a node based on ovirt node 2.6.1. > we get the following error after registration completed successfully > and we try to activate the host via webadmin: > > "Installing Host server4 Starting vdsm" > "Host server4 Installation failed. Unexpected connection termination." > > I attached an excerpt from the engine.log, which shows that an ssh > commands fails, but I don't know why. > > manual ssh connections to the node from the management node work just fine. > > The node has the IP 10.0.1.4 in the log. > > Any help would be appreciated. > > Regards > > Sven > > PS: > The same Node works fine with oVirt management 3.2. > We then did use the "reinstall" feature from the node > iso to redeploy the node to the management 3.3 server. > > > > _______________________________________________ > Users mailing list > Users(a)ovirt.org > http://lists.ovirt.org/mailman/listinfo/users >

----- Original Message ----- You are trying to update not deploy... and you missing the iso image? 2013-10-09 10:50:51,423 ERROR [org.ovirt.engine.core.bll.GetoVirtISOsQuery] (ajp--127.0.0.1-8702-11) ovirt ISOs directory not found. Search in: /usr/share/ovirt-node-iso

Am Donnerstag, den 10.10.2013, 04:05 -0400 schrieb Alon Bar-Lev: Hey Alon, besides this error we also saw this: /var/log/ovirt-engine/engine.log:2013-10-10 08:35:27,016 ERROR [org.ovirt.engine.core.utils.ssh.SSHDialog] (pool-6-thread-4) SSH stderr during command root@10.0.1.4:'umask 0077; MYTMP="$(mktemp -t ovirt-XXXXXXXXXX)"; trap "chmod -R u+rwX \"${MYTMP}\" > /dev/null 2>&1; rm -fr \"${MYTMP}\" > /dev/null 2>&1" 0; rm -fr "${MYTMP}" && mkdir "${MYTMP}" && tar --warning=no-timestamp -C "${MYTMP}" -x && "${MYTMP}"/setup DIALOG/dialect=str:machine DIALOG/customization=bool:True': stderr: bash: line 1: 10803 Segmentation fault "${MYTMP}"/setup DIALOG/dialect=str:machine DIALOG/customization=bool:True which shows that there is some segfault. Is there a way to enable debugging for this bash snippet? E.g. to keep the MYTMP dir if something goes wrong? Or enable bash -x (if setup is a bash script) to see what is being called. To me it's a black box what's happening inside "setup". Greetings fabian

Hi, just in case someone else runs into the same problem. We were able to workaround this by using an alternate draft-iso image kindly provided by fabian: http://fedorapeople.org/~fabiand/node/ovirt-node-iso-3.0.1-1.0.2013100208... if you want the fedora version it's: http://fedorapeople.org/~fabiand/node/ovirt-node-iso-3.1.0-0.999.20131002... We didn't test the fedora version. This iso registers completly with Mangement 3.3 on the same host, no more segfaults :-) A big "thank you" goes out to fabian, who helped me debugging this. Regards Sven