On April 29, 2020 7:42:55 PM GMT+03:00, Shareef Jalloq <[email protected]> wrote: >Ah of course. I was assuming something had gone wrong with the >deployment >and it couldn't clean up its own mess. I'll raise a bug on the >documentation. > >Strahil, what are the other options to using /dev/sdxxx? > >On Wed, Apr 29, 2020 at 10:17 AM Strahil Nikolov ><[email protected]> >wrote: > >> On April 29, 2020 2:39:05 AM GMT+03:00, Jayme <[email protected]> >wrote: >> >Has the drive been used before, it might have existing >> >partition/filesystem >> >on it? If you are sure it's fine to overwrite try running wipefs -a >> >/dev/sdb on all hosts. Also make sure there aren't any filters setup >in >> >lvm.conf (there shouldn't be on fresh install, but worth checking). >> > >> >On Tue, Apr 28, 2020 at 8:22 PM Shareef Jalloq ><[email protected]> >> >wrote: >> > >> >> Hi, >> >> >> >> I'm running the gluster deployment flow and am trying to use a >second >> >> drive as the gluster volume. It's /dev/sdb on each node and I'm >> >using the >> >> JBOD mode. >> >> >> >> I'm seeing the following gluster ansible task fail and a google >> >search >> >> doesn't bring up much. >> >> >> >> TASK [gluster.infra/roles/backend_setup : Create volume groups] >> >> **************** >> >> >> >> failed: [ovirt-gluster-01.jalloq.co.uk] (item={u'vgname': >> >> u'gluster_vg_sdb', u'pvname': u'/dev/sdb'}) => >{"ansible_loop_var": >> >"item", >> >> "changed": false, "err": " Couldn't find device with uuid >> >> Y8FVs8-LP6w-R6CR-Yosh-c40j-17XP-ttP3Np.\n Couldn't find device >with >> >uuid >> >> tA4lpO-hM9f-S8ci-BdPh-lTve-0Rh1-3Bcsfy.\n Couldn't find device >with >> >uuid >> >> RG3w6j-yrxn-2iMw-ngd0-HgMS-i5dP-CGjaRk.\n Couldn't find device >with >> >uuid >> >> lQV02e-TUZE-PXCd-GWEd-eGqe-c2xC-pauHG7.\n Device /dev/sdb >excluded >> >by a >> >> filter.\n", "item": {"pvname": "/dev/sdb", "vgname": >> >"gluster_vg_sdb"}, >> >> "msg": "Creating physical volume '/dev/sdb' failed", "rc": 5} >> >> _______________________________________________ >> >> Users mailing list -- [email protected] >> >> To unsubscribe send an email to [email protected] >> >> Privacy Statement: https://www.ovirt.org/privacy-policy.html >> >> oVirt Code of Conduct: >> >> https://www.ovirt.org/community/about/community-guidelines/ >> >> List Archives: >> >> >> > >> >https://lists.ovirt.org/archives/list/[email protected]/message/5U3K3IPYCFOLUFJ56FGJI3TYWT6NOLAZ/ >> >> >> >> Actually best practice is not to use /dev/sdxxx as they can change. >> >> In your case most peobably the LUN is not fresh, so wipe it with >> dd/blktrim so any remnants of old FS signature is gone. >> >> Best Regards, >> Strahil Nikolov >>
Hi Schareef, In general we should use persistent names like '/dev/disk/by-id/scsi-XYZ or /dev/disk/by-id/wwn-XYZ if we want to be idempotent (be able to rerun the ansible play/role multiple times /even after a reboot/). For example , I make an array of all available disks with the exclusion of the system disk (filtering disks with partitions) and then use those as PVs in a VG and then everything else is easy. If you need to separate the disks by size (multiple VGs), you can sort the array and then select which disk to be a PV for a specific VG. Aanother approach is to filter the disks by vendor or type and then create your VGs with ansible. Anyway, for initial deployment /dev/sdXYZ is enough. Best Regards, Strahil Nikolov _______________________________________________ Users mailing list -- [email protected] To unsubscribe send an email to [email protected] Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/[email protected]/message/TJVX2JN74W2UBAE2TRPQ7T2CP2AUXSBV/
