public inbox for kdevops@lists.linux.dev
 help / color / mirror / Atom feed
From: Chuck Lever <cel@kernel.org>
To: <kdevops@lists.linux.dev>
Cc: Luis Chamberlain <mcgrof@kernel.org>,
	Chuck Lever <chuck.lever@oracle.com>
Subject: [PATCH v2 0/8] nfs: few fixes and enhancements
Date: Fri,  3 Oct 2025 16:19:48 -0400	[thread overview]
Message-ID: <20251003201958.24452-1-cel@kernel.org> (raw)

From: Chuck Lever <chuck.lever@oracle.com>

Original cover:

In prepration for talking about NFS tests at the MSST conference today
I figured I'd give a run to all NFS tests. I ran out of time but at
least this plumbed quite a bit of the stuff to get some results out.

The iSCSI stuff is likely not correct, and can be dropped. So, feel
free to take in only what makes sense and drop whatever silly thing
you see.

Updates:

I'm reposting because, strangely, I never received these patches in
my inbox. I pulled this series from lore using "b4 am" so we can
keep reviewing.

For "devconfig: exclude nfsd from journal upload client
configuration", I wonder if instead of "nfsd", the new checks
should look for the "service" group, which usually includes nfsd,
the SMB server, the iscsi target, and the Kerberos KDC. Any opinion
on that?

I've dropped the iSCSI-specific changes. Except for a couple of
nits, the remaining patches look great to me.

The original idea for pNFS block testing was that a separate iSCSI
target was to be set up: either it is outside the kdevops test
network, or it is enabled by setting CONFIG_KDEVOPS_ENABLE_ISCSI.
Then change the kdevops nfsd server to use that iSCSI target by
changing the "Persistent storage for exported file systems" setting
to "iSCSI".

It's easier overall if the iSCSI target host is separate from the
kdevops nfsd host. iSCSI loopback is not as performant and is less
reliable, I've found -- and maybe not supported on every Linux
distribution we want to run kdevops on, IIRC.

The gitr, nfstest, and pynfs workflows should all be able to use
pNFS block, and indeed the latter two have additional tests
especially for pNFS block layout. So fstests-specific pNFS block
patches don't seem right to me.

Luis, if you believe I missed something and need to revisit one or
more of the patches I've left out, please don't hesitate to bring
it up.

Luis Chamberlain (8):
  defconfigs: add NFS testing configurations
  devconfig: exclude nfsd from journal upload client configuration
  iscsi: add missing initiator packages for Debian
  nfsd_add_export: fix become method for filesystem formatting
  workflows: fstests: fix incorrect pNFS export configuration
  nfstest: add results visualization support
  fstests: add soak duration to nfs template
  pynfs: add visualization support for test results

 defconfigs/nfs-fstests                        |   38 +
 defconfigs/nfs-gitr                           |   38 +
 defconfigs/nfs-ltp                            |   31 +
 defconfigs/nfstests                           |   30 +
 defconfigs/pynfs-pnfs-block                   |   34 +
 playbooks/roles/devconfig/tasks/main.yml      |    3 +
 playbooks/roles/fstests/tasks/main.yml        |    5 +-
 .../roles/fstests/templates/nfs/nfs.config    |    4 +
 playbooks/roles/iscsi/vars/Debian.yml         |    3 +
 .../nfsd_add_export/tasks/storage/local.yml   |    3 +-
 scripts/workflows/pynfs/visualize_results.py  | 1014 +++++++++++++++++
 workflows/Makefile                            |    4 +
 workflows/nfstest/Makefile                    |    1 +
 .../nfstest/scripts/generate_nfstest_html.py  |  783 +++++++++++++
 .../nfstest/scripts/parse_nfstest_results.py  |  277 +++++
 .../scripts/visualize_nfstest_results.sh      |   61 +
 workflows/pynfs/Makefile                      |   17 +-
 17 files changed, 2341 insertions(+), 5 deletions(-)
 create mode 100644 defconfigs/nfs-fstests
 create mode 100644 defconfigs/nfs-gitr
 create mode 100644 defconfigs/nfs-ltp
 create mode 100644 defconfigs/nfstests
 create mode 100644 defconfigs/pynfs-pnfs-block
 create mode 100755 scripts/workflows/pynfs/visualize_results.py
 create mode 100755 workflows/nfstest/scripts/generate_nfstest_html.py
 create mode 100755 workflows/nfstest/scripts/parse_nfstest_results.py
 create mode 100755 workflows/nfstest/scripts/visualize_nfstest_results.sh

-- 
2.51.0


             reply	other threads:[~2025-10-03 20:20 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-03 20:19 Chuck Lever [this message]
2025-10-03 20:19 ` [PATCH v2 1/8] defconfigs: add NFS testing configurations Chuck Lever
2025-10-03 20:19 ` [PATCH v2 2/8] devconfig: exclude nfsd from journal upload client configuration Chuck Lever
2025-10-03 20:19 ` [PATCH v2 3/8] iscsi: add missing initiator packages for Debian Chuck Lever
2025-10-03 20:19 ` [PATCH v2 4/8] nfsd_add_export: fix become method for filesystem formatting Chuck Lever
2025-10-03 20:19 ` [PATCH v2 5/8] workflows: fstests: fix incorrect pNFS export configuration Chuck Lever
2025-10-03 20:19 ` [PATCH v2 6/8] nfstest: add results visualization support Chuck Lever
2025-10-03 20:19 ` [PATCH v2 7/8] fstests: add soak duration to nfs template Chuck Lever
2025-10-03 20:19 ` [PATCH v2 8/8] pynfs: add visualization support for test results Chuck Lever
2025-10-03 22:57 ` [PATCH v2 0/8] nfs: few fixes and enhancements Luis Chamberlain

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=20251003201958.24452-1-cel@kernel.org \
    --to=cel@kernel.org \
    --cc=chuck.lever@oracle.com \
    --cc=kdevops@lists.linux.dev \
    --cc=mcgrof@kernel.org \
    /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