[pve-devel] [PATCH qemu-server] qmp client: bump timeout for synchronous blockdev snapshot commands to 1 hour

[pve-devel] [PATCH qemu-server] qmp client: bump timeout for synchronous blockdev snapshot commands to 1 hour

[Fiona Ebner]

It's a commonly reported issue, most recently again in the enterprise
support, that taking or removing snapshots of large qcow2 files on
file-based network storages runs into a timeout. If the timeout is
hit, that just means additional manual steps to clean up afterwards
for users. The synchronous QMP command will still be blocking the main
thread of the QEMU instance until it has run to completion. Therefore,
a longer timeout does not really hurt here, while reducing the number
of times the aforementioned cleanup is necessary.

In the long term, using snapshot-as-volume-chain should be a good
alternative for such cases, once it's deemed production-ready.

Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
 src/PVE/QMPClient.pm | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/src/PVE/QMPClient.pm b/src/PVE/QMPClient.pm
index e4f7a515..c3ed0e32 100644
--- a/src/PVE/QMPClient.pm
+++ b/src/PVE/QMPClient.pm
@@ -132,8 +132,6 @@ sub cmd {
             $cmd->{execute} eq 'backup-cancel'
             || $cmd->{execute} eq 'blockdev-del'
             || $cmd->{execute} eq 'blockdev-mirror'
-            || $cmd->{execute} eq 'blockdev-snapshot-delete-internal-sync'
-            || $cmd->{execute} eq 'blockdev-snapshot-internal-sync'
             || $cmd->{execute} eq 'block-job-cancel'
             || $cmd->{execute} eq 'block-job-complete'
             || $cmd->{execute} eq 'drive-mirror'
@@ -146,6 +144,11 @@ sub cmd {
             || $cmd->{execute} eq 'savevm-start'
         ) {
             $timeout = 10 * 60; # 10 mins
+        } elsif (
+            $cmd->{execute} eq 'blockdev-snapshot-delete-internal-sync'
+            || $cmd->{execute} eq 'blockdev-snapshot-internal-sync'
+        ) {
+            $timeout = 60 * 60; # 1 hour
         } else {
             #  NOTE: if you came here as user and want to change this, try using IO-Threads first
             # which move out quite some processing of the main thread, leaving more time for QMP
-- 
2.47.3



_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel

[pve-devel] applied: [PATCH qemu-server] qmp client: bump timeout for synchronous blockdev snapshot commands to 1 hour

[Thomas Lamprecht]

On Fri, 28 Nov 2025 15:56:11 +0100, Fiona Ebner wrote:
> It's a commonly reported issue, most recently again in the enterprise
> support, that taking or removing snapshots of large qcow2 files on
> file-based network storages runs into a timeout. If the timeout is
> hit, that just means additional manual steps to clean up afterwards
> for users. The synchronous QMP command will still be blocking the main
> thread of the QEMU instance until it has run to completion. Therefore,
> a longer timeout does not really hurt here, while reducing the number
> of times the aforementioned cleanup is necessary.
> 
> [...]

Applied, thanks!

[1/1] qmp client: bump timeout for synchronous blockdev snapshot commands to 1 hour
      commit: e5f7156b1d2a0c2a1d8705924aed1c22b7fa210f


_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel