Bug 1656520
Summary: | NetworkManager keeps the activated status on DHCP renew failure | ||||||
---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 7 | Reporter: | Matteo Croce <mcroce> | ||||
Component: | NetworkManager | Assignee: | Beniamino Galvani <bgalvani> | ||||
Status: | CLOSED INSUFFICIENT_DATA | QA Contact: | Desktop QE <desktop-qa-list> | ||||
Severity: | medium | Docs Contact: | |||||
Priority: | medium | ||||||
Version: | 7.6 | CC: | atragler, bgalvani, fgiudici, lrintel, pasik, rkhan, sukulkar, thaller | ||||
Target Milestone: | rc | ||||||
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: | 2020-04-23 07:28:25 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
Matteo Croce
2018-12-05 16:58:57 UTC
I have discovered what triggered the bug. The DHCP server was not responding because the IP and subnet of the DHCP server changed between the offer and the renew. So NetworkManager was still sending the renew to the old IP address. Bringing the connection down and up manually restored the connectivity. Created attachment 1511834 [details]
journalctl -b -u NetworkManager
Hi, at 'Dec 04 21:05:47' interface em3 gets a lease with 1 day lifetime: dhcp4 (em3): address 10.19.188.157 dhcp4 (em3): plen 23 (255.255.254.0) dhcp4 (em3): gateway 10.19.189.254 dhcp4 (em3): lease time 86400 and dhclient schedules the renewal in ~9 hours: dhclient[11256]: bound to 10.19.188.157 -- renewal in 33801 seconds. At 'Dec 05 06:29:08' dhclient starts trying to renew the lease without success: dhclient[11256]: DHCPREQUEST on em3 to 10.16.36.29 port 67 (xid=0x147aa2dc) dhclient[11256]: DHCPREQUEST on em3 to 10.16.36.29 port 67 (xid=0x147aa2dc) ... and keeps doing so until it is able to get a new lease at 'Dec 05 18:05:53': dhclient[11256]: DHCPREQUEST on em3 to 255.255.255.255 port 67 (xid=0x147aa2dc) dhclient[11256]: DHCPACK from 10.19.189.254 (xid=0x147aa2dc) During this interval the old lease is still valid and so, even this can't be seen from the log, em3 should still have an address. After one minute, at 'Dec 05 18:06:58', the connection is manually restarted. Do you remember at which time you ran 'ip a' and saw there was no address? Hi, I checked the connection via IPMI console at 17:15, and the lease obtained was because I manually ran 'dhclient em3'. Before running dhclient, with tcpdump I saw DHCP requests going out, so I tought that the DHCP server was down and run dhclient manually. Hi, from the logs attached it is not possible to understand what the problem is. According to logs, the lease was renewed before the expiry time and so the IPv4 address should have always been there. I don't know why it was removed. I'm closing this bz for now. Please reopen if you are able to reproduce the problem; in such case, please also increase NM logging level to TRACE and attach the log. Thanks! |