seems NM may send requests to delete qdisc, which kernel replies with
> Mar 22 10:32:02.293988 dev-h-vrt-013 NetworkManager[11102]: <debug> [1584865922.2939] platform: (ens1f0) qdisc: delete fq_codel dev 865 family 0 handle 0 parent 1 info 1 limit 10240 flows 1024 target 4999 interval 99999 quantum 1514 memory_limit 33554432 ecn
> Mar 22 10:32:02.294009 dev-h-vrt-013 NetworkManager[11102]: <debug> [1584865922.2940] platform-linux: netlink: recvmsg: error message from kernel: No such file or directory (2) "Failed to find qdisc with specified classid" for request 604864
> Mar 22 10:32:02.294018 dev-h-vrt-013 NetworkManager[11102]: <debug> [1584865922.2940] platform-linux: do-delete-qdisc[865: 1]: failure 2 (No such file or directory - Failed to find qdisc with specified classid), meaning the object was already removed
NM then thinks that the element is already gone and that its cache is out of date. It will thus request a full dump of the QDiscs to fix the (assumed) cache consistency.
that means, if you try to delete N qdiscs, not only will it fail, it will also result in N full dumps of all qdiscs. That is a scalability issue as well.
See log here: https://bugzilla.redhat.com/show_bug.cgi?id=1815060#c12
Comment 3Beniamino Galvani
2020-05-25 13:33:16 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory (NetworkManager bug fix and enhancement update), and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.
https://access.redhat.com/errata/RHBA-2020:4499