From: Thomas Lamprecht <t.lamprecht@proxmox.com>
To: pve-devel@lists.proxmox.com
Subject: Re: [PATCH v8 cluster,storage,docs,manager 00/43] SAN: cluster-wide multipath, SAN views and a guided setup for shared LVM
Date: Fri, 20 Nov 2026 09:45:20 +0100 [thread overview]
Message-ID: <179516247284.2632472.8451368514027474084.b4-ty@b4> (raw)
In-Reply-To: <20261119194655.3163258-1-t.lamprecht@proxmox.com>
On Thu, 19 Nov 2026 20:46:05 +0100, Thomas Lamprecht wrote:
> This is v8 of the combined and reduced series of the cluster-wide
> multipath work [0][1] and Dietmar's RFC for guided remote storage setup
> and SAN visibility [2].
>
> If nothing grave comes up I plan to apply this as is.
>
> v8 is v7 [9] plus what an independent review of v7 turned up and held
> up when checked against the code and the lab, i.e.:
>
> [...]
Applied, big thanks here to Blockbridge [0], which where very reactive
and tested basically every version here, ensuring a much better end
result!
[0]: https://www.blockbridge.com
Compared to v8 as posted, I folded in fixes for quite a few more findings,
but mostly rather targeted edge cases:
- If the multipath map of a LUN was not set up on a node, for example
because multipathd failed there, LVM on that node hid the PV on the
LUN, so a wipe from there could destroy the volume group of a shared
storage. The checks before a wipe or a volume group creation now see
such PVs, and also read the LVM label directly.
- The wipe and the volume group creation checked the LUN through its
name, which can lead to another device by the time they write. Both
now check the device they hold open.
- Creating a volume group through the disk API on a LUN that already
had one of the same name, for example from another node, just
succeeded, so two guided setups on one LUN could end with two storages
on the same volume group. This is refused now; only adding a storage
with the LUN as base volume takes over an existing volume group.
- A wipe of a multipath map without any working path could hang in its
checks until the cluster lock timed out. It is now refused right away,
as the volume group creation already did.
- An entry with the name of a map instead of a WWID was accepted by
'multipath -a' without an error, but never showed up as added, so the
node reloaded multipathd on every pass. Such an entry is now reported
as refused.
- The status report that each node publishes only had room for about
180 LUNs. It is stored compressed now, which raises that to about 700.
- The LUN list, the checks and the status each worked out on their own
which LUN an LVM storage uses as base, and did not always agree: a
base with a partition suffix, for example, did not protect its LUN
from a wipe. They now use one helper.
- The cluster status still expected the LUN of a disabled LVM storage
on its nodes, so the LUN showed as missing everywhere once its iSCSI
sessions were gone. A disabled storage now expects nothing.
- A hand-edited overrides file skipped the syntax check that the API
does, so it could hold a multipaths section that the API refuses.
Such a file now stops the apply with an error, instead of being
installed.
- A node without iscsiadm, but in the scope of an iSCSI storage with
multipath management, could neither apply its Fibre Channel entries
nor show Node > SAN. It now only asks iscsiadm about targets that
have paths on that node.
- Adding a second entry for the same WWID now shows the error at the
WWID field, instead of as a generic error.
- If the base volume of an LVM storage could not be activated, a status
pass repeated slow LVM scans for it. It now stops at the first
failure.
- In the web UI, a LUN that is unknown on all nodes no longer lists
every node in its state, the Target/Subsystem header is one string
for translators, and the iSCSI setup says where to allow the
initiator names.
- The multipath chapter is shorter and simpler, and also covers the
options of a multipath entry and that nothing notifies about a lost
path. The web UI and pvestatd chapters mention the SAN panels and
what pvestatd does for multipath.
- Plus some rewording of comments, test descriptions and commit
messages.
Still open (just for sake of completness):
- Nothing notifies about a lost path yet, so an admin has to look at
the SAN view. A notification could build on the status that every
node already reports.
- Storage plugins cannot tell how many paths a volume should have, so
a namespace that a plugin connected looks single-path once one of its
paths is gone. A plugin method for the expected paths could add that.
- NVMe over Fabrics support follows as v3 of its own series, on top of
this.
[pve-cluster]
[01/43] cfs: observe the multipath configuration files
commit: f196907143173a904a24a42f7c2a9f7f78c183c4
[pve-storage]
[02/43] diskmanage: collect disk transport type from lsblk
commit: d5167f815b6344718506e4a5449096ed9518bf4c
[03/43] diskmanage: qualify NVMe over Fabrics transport
commit: 968f1549f8acc3874d253de8a476664438a18d16
[04/43] diskmanage: optionally include iSCSI session devices in disk enumeration
commit: 08c04c8d7a077ff03f4a6fbdab466037ed6e85dd
[05/43] diskmanage: collect native NVMe multipath path state
commit: 6e52aaf06509dc8314030793e1b76f785ee7367a
[06/43] iscsi: factor out session device map from device list
commit: e9e65e672d19cc5d1f33c13a5175cb1eee90fd34
[07/43] multipath: add a library around multipathd
commit: dfe34094562cfc5de060ff835b6e75debc2be12a
[08/43] diskmanage: add helper to list multipath devices
commit: d3e5f28dae60dac121a46a42d63527e5db6f03b5
[09/43] multipath: add a cluster-wide configuration
commit: 59948403ba368b3b771f4860f359c62d1347bd3e
[10/43] iscsi: add an option to manage storage LUNs with multipath
commit: 55dd1de9a1d8017b23ffe53c2815ce5fcfc5a217
[11/43] lvm: activate a missing base volume before using the volume group
commit: 44bc0e3e4e4b261fc4cc76e47dc233f0b4768694
[12/43] multipath: apply the configuration on every node
commit: 51b8a5c820b9ba68d511022350eab4e9cc614937
[13/43] multipath: publish per-node status
commit: e72b1be2793c36e6991d87fe7707ba5288d05509
[14/43] api: add endpoints for the cluster-wide multipath configuration
commit: f0d87e448c207bf166fd000bb880c35c551bfba5
[15/43] api: multipath: add the cluster-wide health status
commit: c98ecf288ff4423d5064e51d19e28734b99a6040
[16/43] diskmanage: look up the requested disks by name
commit: 8222e318de382ea22f554998dd876b67e8c08ee0
[17/43] diskmanage: read the udev data of all disks with one call
commit: 298dcfaec978ae611e6c54d155bd1cdd80e6d7fb
[18/43] api: scan: add a SAN LUN inventory
commit: 2e5f0afad07ffd1845c7d0940e9662df48c72836
[19/43] pvesm: expose the SAN LUN inventory
commit: 73091d3621e51ad5fccb316d2efd902c0570212d
[20/43] api: add node-level iSCSI status
commit: cdf8763f784d2eddadfacaa576d39b64b0c8dea4
[21/43] api: scan: name the portal of an iSCSI path
commit: ad2ccb2ead41c6f4675d9a8d9e382dd59863d7f6
[22/43] disks: lvm: drop unused imports
commit: 946c8422ed805512515864fe80b6aaae340315ec
[23/43] disks: lvm: allow volume groups on SAN LUNs and multipath maps
commit: 6a98c4ba61fb133a6dc82bf20931417f5c0c5ed8
[24/43] disks: lvm: refuse to destroy the volume group of a shared storage
commit: 679d241b7bb4c016704b2fd6ed7983a4983efd98
[25/43] disks: allow wiping a SAN LUN that no storage uses
commit: c5529caa6377f03654d121573659e1cfc0235074
[26/43] api: scan: report the process that has a SAN LUN open
commit: 8e9935f595a664290e6edd4ae8c23b28fe631018
[pve-docs]
[27/43] storage: document the cluster-wide multipath management
commit: 5704d152297524600925469f2065e66d21d49300
[28/43] storage: multipath: describe the use shown for an open LUN
commit: 64861a0de0a5e086f3f5557a6e9165cf75ff8996
[pve-manager]
[29/43] pvestatd: apply the multipath configuration and report the node status
commit: ec55d57a3288630c0690dea4e8d63280e70d58c3
[30/43] report: add the managed multipath configuration
commit: a87eab65140ac4ae098ee4304830d3b9815565a1
[31/43] api: cluster: add the multipath endpoints
commit: 0a6f615a5594135103c3f6aaedeaeefb08664114
[32/43] api: nodes: expose the iSCSI status
commit: faaaadcf6f29c7cd7b1d938d88f0957ceba4d5f8
[33/43] ui: dc: add a SAN view
commit: a335adb38a9b2b9d0a05a1e3afc720bf9a0897c1
[34/43] ui: storage: iscsi: add the multipath option
commit: ca0bfb7e4b079e328f60037ba784ec2354c836b4
[35/43] ui: lvm: enable shared when a base storage is selected
commit: 6cfb3b8ac945ae4b7b0b07ac122da7c5d5f20584
[36/43] ui: node: add SAN LUNs panel
commit: d079fb4b3bd3126df29684822083922e2469afde
[37/43] ui: node: san: wipe a LUN that no storage uses
commit: 966f798adacd34d5b726ae9e6265a8a02a7ec5c5
[38/43] ui: node: san: show iSCSI targets and sessions
commit: 56482f5dd485d37a06f33329216c375e135d5b33
[39/43] ui: dc: san: add a guided setup for shared LVM on iSCSI
commit: e2094f57fff7b8058060b0fb7d0410c2ab3ec744
[40/43] ui: dc: san: add a guided setup for shared LVM on Fibre Channel
commit: 39162712816a5fc715d00a07d86f683b9a005bf9
[41/43] ui: storage: point the iSCSI and LVM add dialogs to the SAN view
commit: ab30edfcf78d6b6017ebb82354a20402dab22d3b
[42/43] ui: san: show the guest or process that has a LUN open
commit: 188f844baad0ce05b701f0f1a9018ff05f207a25
[43/43] ui: node: san: add a column for the target or subsystem of a LUN
commit: 215871e61ea8ab696d5a4400707d2ea18fb46fa2
prev parent reply other threads:[~2026-11-20 8:48 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-19 19:46 [PATCH v8 cluster,storage,docs,manager 00/43] SAN: cluster-wide multipath, SAN views and a guided setup for shared LVM Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 cluster 01/43] cfs: observe the multipath configuration files Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 02/43] diskmanage: collect disk transport type from lsblk Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 03/43] diskmanage: qualify NVMe over Fabrics transport Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 04/43] diskmanage: optionally include iSCSI session devices in disk enumeration Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 05/43] diskmanage: collect native NVMe multipath path state Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 06/43] iscsi: factor out session device map from device list Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 07/43] multipath: add a library around multipathd Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 08/43] diskmanage: add helper to list multipath devices Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 09/43] multipath: add a cluster-wide configuration Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 10/43] iscsi: add an option to manage storage LUNs with multipath Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 11/43] lvm: activate a missing base volume before using the volume group Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 12/43] multipath: apply the configuration on every node Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 13/43] multipath: publish per-node status Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 14/43] api: add endpoints for the cluster-wide multipath configuration Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 15/43] api: multipath: add the cluster-wide health status Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 16/43] diskmanage: look up the requested disks by name Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 17/43] diskmanage: read the udev data of all disks with one call Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 18/43] api: scan: add a SAN LUN inventory Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 19/43] pvesm: expose the " Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 20/43] api: add node-level iSCSI status Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 21/43] api: scan: name the portal of an iSCSI path Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 22/43] disks: lvm: drop unused imports Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 23/43] disks: lvm: allow volume groups on SAN LUNs and multipath maps Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 24/43] disks: lvm: refuse to destroy the volume group of a shared storage Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 25/43] disks: allow wiping a SAN LUN that no storage uses Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 storage 26/43] api: scan: report the process that has a SAN LUN open Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 docs 27/43] storage: document the cluster-wide multipath management Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 docs 28/43] storage: multipath: describe the use shown for an open LUN Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 29/43] pvestatd: apply the multipath configuration and report the node status Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 30/43] report: add the managed multipath configuration Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 31/43] api: cluster: add the multipath endpoints Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 32/43] api: nodes: expose the iSCSI status Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 33/43] ui: dc: add a SAN view Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 34/43] ui: storage: iscsi: add the multipath option Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 35/43] ui: lvm: enable shared when a base storage is selected Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 36/43] ui: node: add SAN LUNs panel Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 37/43] ui: node: san: wipe a LUN that no storage uses Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 38/43] ui: node: san: show iSCSI targets and sessions Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 39/43] ui: dc: san: add a guided setup for shared LVM on iSCSI Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 40/43] ui: dc: san: add a guided setup for shared LVM on Fibre Channel Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 41/43] ui: storage: point the iSCSI and LVM add dialogs to the SAN view Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 42/43] ui: san: show the guest or process that has a LUN open Thomas Lamprecht
2026-11-19 19:46 ` [PATCH v8 manager 43/43] ui: node: san: add a column for the target or subsystem of a LUN Thomas Lamprecht
2026-11-20 8:45 ` Thomas Lamprecht [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=179516247284.2632472.8451368514027474084.b4-ty@b4 \
--to=t.lamprecht@proxmox.com \
--cc=pve-devel@lists.proxmox.com \
/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