From: Chuck Lever <chuck.lever@oracle.com>
To: Jeff Layton <jlayton@kernel.org>
Cc: kdevops@lists.linux.dev
Subject: Re: [PATCH v2] nfsd: rework how storage is allocated to the LVM VG
Date: Sun, 3 Mar 2024 13:19:35 -0500 [thread overview]
Message-ID: <ZeS_NzZQoR_dfRS0@manet.1015granger.net> (raw)
In-Reply-To: <20240301193156.91590-1-jlayton@kernel.org>
On Fri, Mar 01, 2024 at 02:31:56PM -0500, Jeff Layton wrote:
> Currently we require the user to put in two different path strings, and
> they have to reset that if the configuration changes. This is limiting,
> since we need to add the ability to suck in a 3rd disk.
>
> Rework it to take a device prefix string and a count. "count" number of
> devices will be added to the VG, starting at index 1.
>
> Signed-off-by: Jeff Layton <jlayton@kernel.org>
No objection to any of these. My current kdevops system is too small
to run more than three guests at once, so I haven't hit this
particular limitation so far. My new system will be able to, though.
> ---
> kconfigs/Kconfig.nfsd | 26 ++++++++++++--------------
> playbooks/roles/nfsd/defaults/main.yml | 3 +++
> playbooks/roles/nfsd/tasks/main.yml | 11 ++++++++++-
> scripts/nfsd.Makefile | 4 ++--
> 4 files changed, 27 insertions(+), 17 deletions(-)
>
> v2: I ended up sending an earlier draft by accident. This one actually
> works.
>
> diff --git a/kconfigs/Kconfig.nfsd b/kconfigs/Kconfig.nfsd
> index 946c2c101b58..db80d8c0981e 100644
> --- a/kconfigs/Kconfig.nfsd
> +++ b/kconfigs/Kconfig.nfsd
> @@ -1,22 +1,20 @@
> # FIXME: need defaults for terraform hosts
> -config NFSD_EXPORT_DEVICE_0
> - string "First device to be exported"
> - default "/dev/disk/by-id/nvme-QEMU_NVMe_Ctrl_kdevops2" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_NVME
> - default "/dev/disk/by-id/virtio-kdevops2" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_VIRTIO
> - default "/dev/disk/by-id/ata-QEMU_HARDDISK_kdevops2" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_IDE
> +config NFSD_EXPORT_DEVICE_PREFIX
> + string "The device prefix to use for LVM PVs"
> + default "/dev/disk/by-id/nvme-QEMU_NVMe_Ctrl_kdevops" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_NVME
> + default "/dev/disk/by-id/virtio-kdevops" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_VIRTIO
> + default "/dev/disk/by-id/ata-QEMU_HARDDISK_kdevops" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_IDE
> default ""
> help
> - To set up nfsd for testing, we give it filesystems to export. This
> - block device is used as a PV for LVM, and volumes are provisioned from it.
> + To set up nfsd for testing, we give it filesystems to export. This string
> + will be the prefix for the block devices used as PVs for LVM.
>
> -config NFSD_EXPORT_DEVICE_1
> - string "Second device to be exported"
> - default "/dev/disk/by-id/nvme-QEMU_NVMe_Ctrl_kdevops3" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_NVME
> - default "/dev/disk/by-id/virtio-kdevops3" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_VIRTIO
> - default "/dev/disk/by-id/ata-QEMU_HARDDISK_kdevops3" if LIBVIRT && LIBVIRT_EXTRA_STORAGE_DRIVE_IDE
> +config NFSD_EXPORT_DEVICE_COUNT
> + int "Number of devices to add as LVM PVs"
> + default 3 if LIBVIRT
> help
> - To set up nfsd for testing, we give it filesystems to export. This
> - block device is used as a PV for LVM, and volumes are provisioned from it.
> + The number of disk devices to dedicate as LVM PVs. In general, we
> + avoid using device index 0 as that is used for /data.
>
> choice
> prompt "Type of filesystem to export"
> diff --git a/playbooks/roles/nfsd/defaults/main.yml b/playbooks/roles/nfsd/defaults/main.yml
> index 4a72f4f0b048..38718aec2bf7 100644
> --- a/playbooks/roles/nfsd/defaults/main.yml
> +++ b/playbooks/roles/nfsd/defaults/main.yml
> @@ -1,6 +1,9 @@
> # SPDX-License-Identifier GPL-2.0+
> ---
> # Our sensible defaults for the nfsd role.
> +nfsd_lvm_pvs: []
> +nfsd_export_device_prefix: ""
> +nfsd_export_device_count: 0
> nfsd_export_label: "export"
> nfsd_export_fs_opts: ""
> nfsd_lease_time: "90"
> diff --git a/playbooks/roles/nfsd/tasks/main.yml b/playbooks/roles/nfsd/tasks/main.yml
> index e051d21bd819..fe5f6919864d 100644
> --- a/playbooks/roles/nfsd/tasks/main.yml
> +++ b/playbooks/roles/nfsd/tasks/main.yml
> @@ -22,13 +22,22 @@
> group: root
> mode: 0644
>
> +- name: Build string of devices to use as PVs
> + set_fact:
> + nfsd_lvm_pvs: "{{ nfsd_lvm_pvs + [ nfsd_export_device_prefix + item|string ] }}"
> + with_items: "{{ range(1, nfsd_export_device_count + 1) }}"
> +
> +- name: Print the PV list
> + ansible.builtin.debug:
> + var: nfsd_lvm_pvs
> +
> - name: Create a new LVM VG
> become: yes
> become_flags: 'su - -c'
> become_method: sudo
> community.general.lvg:
> vg: "exports"
> - pvs: "{{ nfsd_export_device_0 }},{{ nfsd_export_device_1 }}"
> + pvs: "{{ nfsd_lvm_pvs | join(',') }}"
>
> - name: Create {{ nfsd_export_path }}
> become: yes
> diff --git a/scripts/nfsd.Makefile b/scripts/nfsd.Makefile
> index 40338ab93fce..d3549a29af84 100644
> --- a/scripts/nfsd.Makefile
> +++ b/scripts/nfsd.Makefile
> @@ -1,5 +1,5 @@
> -NFSD_EXTRA_ARGS += nfsd_export_device_0='$(subst ",,$(CONFIG_NFSD_EXPORT_DEVICE_0))'
> -NFSD_EXTRA_ARGS += nfsd_export_device_1='$(subst ",,$(CONFIG_NFSD_EXPORT_DEVICE_1))'
> +NFSD_EXTRA_ARGS += nfsd_export_device_prefix='$(subst ",,$(CONFIG_NFSD_EXPORT_DEVICE_PREFIX))'
> +NFSD_EXTRA_ARGS += nfsd_export_device_count='$(subst ",,$(CONFIG_NFSD_EXPORT_DEVICE_COUNT))'
> NFSD_EXTRA_ARGS += nfsd_export_fstype='$(subst ",,$(CONFIG_NFSD_EXPORT_FSTYPE))'
> NFSD_EXTRA_ARGS += nfsd_export_path='$(subst ",,$(CONFIG_NFSD_EXPORT_PATH))'
> NFSD_EXTRA_ARGS += nfsd_export_options='$(subst ",,$(CONFIG_NFSD_EXPORT_OPTIONS))'
> --
> 2.43.2
>
>
--
Chuck Lever
prev parent reply other threads:[~2024-03-03 18:19 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-01 19:31 [PATCH v2] nfsd: rework how storage is allocated to the LVM VG Jeff Layton
2024-03-03 18:19 ` Chuck Lever [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZeS_NzZQoR_dfRS0@manet.1015granger.net \
--to=chuck.lever@oracle.com \
--cc=jlayton@kernel.org \
--cc=kdevops@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).