Bug 1360386
| Summary: | bond slaves of master which is slave of a bridge are sometimes not activated in installer | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Radek Vykydal <rvykydal> |
| Component: | NetworkManager | Assignee: | Beniamino Galvani <bgalvani> |
| Status: | CLOSED ERRATA | QA Contact: | Desktop QE <desktop-qa-list> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.2 | CC: | atragler, bgalvani, jkoten, lrintel, rkhan, sukulkar, thaller, vbenes |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | NetworkManager-1.8.0-0.2.git20170215.1d40c5f4.el7 | Doc Type: | If docs needed, set a value |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2017-08-01 09:17:07 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
| Attachments: | |||
|
Description
Radek Vykydal
2016-07-26 14:47:44 UTC
Created attachment 1184294 [details]
NetworkManager log of working case
Sometimes the bond slaves are activated:
NAME UUID TYPE DEVICE..
Bond connection bond0 5da744dd-eb83-43ae-8ecc-f7c6d744b3dd bond bond0...
Bridge connection bridge0 90b95d38-dd04-42e2-a7fa-6be802c84f61 bridge bridge0.
System ens3 67ce3099-826d-486c-a60e-f2fc978a9ec9 802-3-ethernet ens3....
bond0 slave 1 ee1d7778-1a2e-4645-be62-00bd284a2c56 802-3-ethernet ens9....
bond0 slave 2 f7b1f7d7-9f5b-427f-b69a-256985819d2f 802-3-ethernet ens10...
DEVICE TYPE STATE CONNECTION................
bond0 bond connected Bond connection bond0.....
bridge0 bridge connected Bridge connection bridge0.
ens10 ethernet connected bond0 slave 2.............
ens3 ethernet connected System ens3...............
ens9 ethernet connected bond0 slave 1.............
lo loopback unmanaged --........................
Created attachment 1184295 [details]
NetworkManager log of failing case
Sometimes the bond slaves are not activated:
NAME UUID TYPE DEVICE..
Bond connection bond0 4a5c651f-2fb1-45f6-b74b-df3146acde8b bond bond0...
Bridge connection bridge0 7616654a-9768-4f0f-b92f-ed38d82c7a9b bridge bridge0.
System ens3 77645a09-9a3e-4b21-847e-1b32cfab162c 802-3-ethernet ens3....
bond0 slave 1 c144a239-85bb-4682-8222-44354cdc6682 802-3-ethernet --......
bond0 slave 2 e6d8d114-aa73-4224-8e88-f60a251d9b1a 802-3-ethernet --......
DEVICE TYPE STATE CONNECTION................
bond0 bond connected Bond connection bond0.....
bridge0 bridge connected Bridge connection bridge0.
ens10 ethernet connected --........................
ens3 ethernet connected System ens3...............
ens9 ethernet connected --........................
lo loopback unmanaged --........................
nmcli con up on the slaves makes the configuration work.
Created attachment 1184296 [details]
ifcfg files
ifcfg files created for NM by installer in initramfs
Created attachment 1204946 [details]
[PATCH] manager: allow slave connection to auto-connect its slaves
+autoconnect_root_quark (void) use G_DEFINE_QUARK() + _LOGE (LOGD_CORE, + "will NOT activate slave this doesn't seem to warrant a error message. Especially as activation does not fail in general. Is this really an invalid configuration? Maybe even just <info>. Rest lgtm (In reply to Thomas Haller from comment #6) > +autoconnect_root_quark (void) > > use G_DEFINE_QUARK() > > > + _LOGE (LOGD_CORE, > + "will NOT activate slave > > this doesn't seem to warrant a error message. Especially as activation does > not fail in general. Is this really an invalid configuration? Maybe even > just <info>. We only check for auto-connect circular dependencies here and thus a INFO level probably is enough. But I expect the activation to fail anyway later (AFAIU the kernel doesn't allow such configurations). Created attachment 1208158 [details]
[PATCH v2] manager: allow slave connection to auto-connect its slaves
v2: changed log level from error to info.
+static GQuark
+autoconnect_root_quark (void)
+{
+ static GQuark quark = 0;
+
+ if (G_UNLIKELY (quark == 0))
+ quark = g_quark_from_static_string ("autoconnect-root");
+ return quark;
+}
should be replaced by G_DEFINE_QUARK()
Created attachment 1209036 [details]
[PATCH v3] manager: allow slave connection to auto-connect its slaves
Applied to master: https://cgit.freedesktop.org/NetworkManager/NetworkManager/commit/?id=632dd26288601c8564e1363f9bbf2f1f96d2dc53 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, 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/RHSA-2017:2299 |