Description of problem: Long story short (1). Single host self hosted engine test setup. One of the drives in a RAID6 died, somehow taking the host off-line. Replaced the drive, restarted the host, broker and agent refused to start due to corrupted meta data file. With the help of Yedidyah Bar David I managed to clear the meta data file (2) and return the host to normal. (1) Full mailing list thread: https://lists.ovirt.org/archives/list/users@ovirt.org/thread/LG5WEGGIDJIZSRCHSZLY2UUSVFBW6GTN/ (2) Undocumented solution: https://lists.ovirt.org/pipermail/users/2016-April/072676.html
The documentation text flag should only be set after 'doc text' field is provided. Please provide the documentation text and set the flag to '?' again.
Sandro, Please give this a priority. Right now it's has no priority/severity set and the PM score is 0.
Hi Gilboa and Didi, I looked into the manual HE metadata cleanup and I am wondering if it is not a better option to use the existing hosted-engine --clean-metadata verb? ~~~ # hosted-engine --clean_metadata --host-id=52 ~~~ Also, are you truly having 52 hosts to support your single HE VM? I hope not. It is not recommended to use more than few hosts that can provide High Availability for your Hosted Engine VM. It is unnecessary complication of things.
Ok, I see now this newer thread: https://lists.ovirt.org/archives/list/users@ovirt.org/message/DXPXGU4FKWJOJZJM7CKAB7ZJZFR7NJZE/.
As you can see (in the ML thread) --clean_metadata failed. I had to manually clean the meta-data file. - Gilboa
Thanks, yeah, I read that later.