multiple dh-ctlplane dhclient processes and undercloud drops off network
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
tripleo |
Expired
|
Undecided
|
Unassigned |
Bug Description
Second time this has happened -
uptime
01:34:27 up 5 days, 2:34, 2 users, load average: 0.32, 0.50, 0.55
(haven't rebooted since the last occurrence, think its ~24 hours?)
ps fax | grep dhclient:
4895 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
10507 tty1 S+ 0:00 \_ grep --color=auto dhcp
22169 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
23734 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
23106 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
23273 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
cat /run/dhclient.
23273
Once this happens, we have to get networking back by rebooting or ifdown+ifup - os-collect-config will fail as it can't reach the metadata services.
So I do that, killing all the dhclients in the middle, and then run os-collect-config --force --one to get the route back, and after that:
18642 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
18808 ? Ss 0:00 dhclient -1 -v -pf /run/dhclient.
/var/lib/
I'm fairly sure its an ensure-bridge issue, but I'm not sure if its an issue in the current code or my patchset only.
Changed in tripleo: | |
status: | New → Incomplete |
[Expired for tripleo because there has been no activity for 60 days.]