View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000621 | AlmaLinux-10 | General | public | 2026-05-20 21:20 | 2026-06-05 13:20 |
| Reporter | bxk-google | Assigned To | |||
| Priority | high | Severity | feature | Reproducibility | always |
| Status | new | Resolution | open | ||
| Summary | 0000621: Request to backport recent GVE patches for Google Cloud overcommited Gen3+ VMs for AlmaLinux 10 | ||||
| Description | Currently, the maximum queue depth supported for gVNIC on overcommited Gen3+ VMs such as N4 is 1K, whereas the maximum queue depth supported on Gen1/Gen2 VMs is 2K. Customers who are migrating their workloads from N2 to N4 have requested higher queue depth support on N4 VMs. GVE driver changes to support greater than 1K queue depth for overcommitted Gen3+ have been upstreamed to Linux kernel recently: https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=a2f19184014f https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=07993df56091 | ||||
| Tags | No tags attached. | ||||
|
|
A decision on this hasn't been made one way or the other by ALESCo yet, but we're working on test kernel builds with these changes. |
|
|
An AlmaLinux 10 test kernel has been built using these commits. https://build.almalinux.org/build/59807 You can install it as follows: dnf upgrade --repofrompath gvnic-qd-backport,https://build.almalinux.org/pulp/content/builds/AlmaLinux-10-$(arch)-59807-br/ NVR for this test kernel is 6.12.0-211.7.4.el10_2.gcp |
|
|
It seems to act as expected in my initial testing. Before: --- [root@alma10-gvnic-qd-test ~]# curl "http://metadata.google.internal/computeMetadata/v1/instance/machine-type" -H "Metadata-Flavor: Google" projects/527193872801/machineTypes/n4-standard-8 [root@alma10-gvnic-qd-test ~]# uname -a Linux alma10-gvnic-qd-test 6.12.0-211.7.3.el10_2.x86_64 #1 SMP PREEMPT_DYNAMIC Tue May 19 12:46:58 EDT 2026 x86_64 GNU/Linux [root@alma10-gvnic-qd-test ~]# ethtool -g eth0 Ring parameters for eth0: Pre-set maximums: RX: 1024 RX Mini: n/a RX Jumbo: n/a TX: 1024 TX push buff len: n/a HDS thresh: n/a --- After: --- [jonathan@alma10-gvnic-qd-test ~]$ curl "http://metadata.google.internal/computeMetadata/v1/instance/machine-type" -H "Metadata-Flavor: Google" projects/527193872801/machineTypes/n4-standard-8 [jonathan@alma10-gvnic-qd-test ~]$ uname -a Linux alma10-gvnic-qd-test 6.12.0-211.7.4.el10_2.gcp.x86_64 #1 SMP PREEMPT_DYNAMIC Thu Jun 4 15:33:18 EDT 2026 x86_64 GNU/Linux [jonathan@alma10-gvnic-qd-test ~]$ ethtool -g eth0 Ring parameters for eth0: Pre-set maximums: RX: 4096 RX Mini: n/a RX Jumbo: n/a TX: 4096 TX push buff len: n/a HDS thresh: n/a --- I do notice that the RX/TX threshold are raised to 4k, not 2k as you mentioned. Is 4k the desired result or is something wrong? |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2026-05-20 21:20 | bxk-google | New Issue | |
| 2026-06-04 17:48 | jonathan | Note Added: 0001219 | |
| 2026-06-05 13:05 | jonathan | Note Added: 0001220 | |
| 2026-06-05 13:10 | jonathan | Note Edited: 0001220 | |
| 2026-06-05 13:11 | jonathan | Note Edited: 0001220 | |
| 2026-06-05 13:20 | jonathan | Note Added: 0001221 |
