Re: ceph-volume: hard to replace ceps-disk now and not good implemented

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sat, Mar 31, 2018 at 6:35 PM, Robin H. Johnson <robbat2@xxxxxxxxxx> wrote:
> On Sat, Mar 31, 2018 at 08:48:25AM -0400, Alfredo Deza wrote:
>> On Sat, Mar 31, 2018 at 2:31 AM, Ning Yao <zay11022@xxxxxxxxx> wrote:
>> > Hi, Sage
>> >
>> > Although we have  ceph-volume now, ceps-volume cannot work as expected
>> > in most of scenarios. For example:
>> >
>> > when we use:
>> >
>> > ceph-volume lvm prepare --filestore --data /dev/sda --journal /dev/sdl1  or
>> > ceph-volume lvm prepare --bluestore --data /dev/sda --block.db /dev/sdl1
>> >
>> > 1. we need to create a Physical Volume for /dev/sda before it can be
>> > used. and ceph-volume does not help us to do that
>>
>> Not sure what you mean by this. In that scenario, ceph-volume will use
>> /dev/sda and convert it to a single logical volume for "data". This is
>> even
>> how our tests work. If you are hitting errors, please give us some
>> examples on how to replicate.
> http://tracker.ceph.com/issues/23496
> http://tracker.ceph.com/issues/23497
> http://tracker.ceph.com/issues/23498
> Ubuntu Trusty for the filestore case above. You closed them all out :-(.

I closed them because ceph-volume cannot work without systemd.

>
> Trusty ships an older blkid, so you cannot use it to get PARTUUID as
> easily. It will print PART_ENTRY_UUID if you pass just a SINGLE
> partition to blkid -p, which has the same content as PARTUUID.

Right, and it can't work with this version of blkid, although a
significant effort could be done to support this case. But then again,
if there is no systemd I don't see how we could put the effort towards
supporting the older blkid

The systemd scenario though is tightly coupled, we rely on it 100%.

>
> In the one bug you stated 'ceph-volume does not support trusty because
> it relies on systemd which is not in Trusty'.
>
> I don't mind having to use Upstart to bring up the OSDs still, but I
> want to have consistent creation of OSDs as I upgrade the cluster from
> Trusty to Xenial/Bionic soon.
>
> ceph-disk is also falling over in cases that previously worked in Jewel,
> which is necessitating the push to ceph-volume already.
>
> I think getting ceph-volume to work with Trusty is a good investment in
> this case, at least for creating the OSDs.

I don't know how Upstart would work in this case. In the systemd case,
we are able to have a high confidence in the state of the machine
as OSDs are coming up aside from being able to persist some details
about the OSD (id, fsid) in the service name.
>
> --
> Robin Hugh Johnson
> Gentoo Linux: Dev, Infra Lead, Foundation Treasurer
> E-Mail   : robbat2@xxxxxxxxxx
> GnuPG FP : 11ACBA4F 4778E3F6 E4EDF38E B27B944E 34884E85
> GnuPG FP : 7D0B3CEB E9B85B1F 825BCECF EE05E6F6 A48F6136
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [CEPH Users]     [Ceph Large]     [Information on CEPH]     [Linux BTRFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]
  Powered by Linux