Bug 2056362
| Summary: | Changing NIC type rolls back the checkpoint | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 8 | Reporter: | peyu | ||||
| Component: | cockpit | Assignee: | Martin Pitt <mpitt> | ||||
| Status: | CLOSED DUPLICATE | QA Contact: | peyu | ||||
| Severity: | high | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 8.6 | CC: | arachman, cshao, lsvaty, lveyde, mavital, peyu, qiyuan, sanja, sbonazzo, weiwang, yaniwang | ||||
| Target Milestone: | rc | Flags: | pm-rhel:
mirror+
|
||||
| Target Release: | --- | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2022-02-23 14:18:38 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | Node | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
|
Description
peyu
2022-02-21 05:07:49 UTC
Configuring a type other than the default type for the NIC's IPv4 and IPv6 will also fail. Created attachment 1862565 [details] page after changing IPv{4,6} to manual Hello peyu, thanks for your report! I tried to reproduce this. > 5. Configure the NIC with IPv4 and IPv6 with "Manual" type and fill in the IP address > 6. Click the "Change the settings" button in the "Connection will be lost" dialog that pops up I changed an "inactive" eth card to IPv4 "manual" with 1.2.3.4/24 and gw 1.2.3.1, then clicked "change the settings". I then changed the IPv6 settings of the same card to "manual" with "ab12::2/32" and gw ab12::1, again clicked "change settings. (These are completely bogus addresses of course) > 7. Check the type of IPv4 and IPv6 This worked fine: IPv4: Address 1.2.3.4/24 via 1.2.3.1 IPv6: Address ab12:0:0:0:0:0:0:2/32 via ab12:0:0:0:0:0:0:1 (See attached screenshot) This was kept even after a page reload, and `nmcli con show 'System eth1'` agrees: ipv4.method: manual ipv4.dns: -- ipv4.addresses: 1.2.3.4/24 ipv4.gateway: 1.2.3.1 ipv6.method: manual ipv6.dns: -- ipv6.addresses: ab12::2/32 ipv6.gateway: ab12::1 > 8. Switch the settings from "Manual" type back to "Automatic"/or other type I switched back to "Automatic", first with keeping the additional manual address. Then again with dropping it, so that both are only "Automatic". I also tried "Link local" and "Disabled", both worked. I use the same version as you (cockpit-system-261-1.el8.noarch). So there must be something subtly different. Most probably NetworkManager runs into some error and does not accept the new settings. Can you please try this again, and capture journalctl --since '10 minutes ago' > /tmp/journal.txt after the issue happens, and attach /tmp/journal.txt here? It's also worth opening the developer console (Ctrl+Shift+J) and watch out for errors/warnings when that happens. Thanks! Thanks! This confirms that it's an unintended checkpoint rollback. I.e. the same root cause as bug #2056386, just that this is harder for me to reproduce. Marius is working on the other one, and this is almost surely going to be addressed by the same fix. Marking this as a duplicate for now, as the other bug has a bit more state. *** This bug has been marked as a duplicate of bug 2056386 *** |