Actions
Bug #16927
open
Multiple ``updaterrd.sh`` processes
Start date:
Due date:
% Done:
0%
Estimated time:
Plus Target Version:
26.07
Release Notes:
Default
Affected Version:
2.8.1
Affected Architecture:
All
Description
In some situations pfSense can end up running the updaterrd script multiple times. This appears to be due to failing to kill previously running processes after making a change requiring it's restart.
[2.8.1-RELEASE][admin@pfsense.box]/root: ps -auxwwd | grep rrd root 10795 0.0 0.0 14644 3248 - IN 25May26 20:59.48 |-- /bin/sh /var/db/rrd/updaterrd.sh root 17980 0.0 0.0 14644 3240 - IN Fri16 3:01.14 |-- /bin/sh /var/db/rrd/updaterrd.sh root 26275 0.0 0.0 14644 3256 - IN Sun19 1:47.85 |-- /bin/sh /var/db/rrd/updaterrd.sh root 28427 0.0 0.0 14644 3236 - IN Sun19 1:46.10 |-- /bin/sh /var/db/rrd/updaterrd.sh root 43465 0.0 0.0 14644 3248 - IN 31Dec25 95:17.41 |-- /bin/sh /var/db/rrd/updaterrd.sh root 44750 0.0 0.0 14644 3248 - SN 30May26 18:00.91 |-- /bin/sh /var/db/rrd/updaterrd.sh root 27660 0.0 0.0 14076 2684 0 S+ 23:43 0:00.00 | `-- grep rrd root 62979 0.0 0.0 14644 3240 - IN 30May26 17:45.98 |-- /bin/sh /var/db/rrd/updaterrd.sh root 67321 0.0 0.0 14644 3240 - IN Sun19 1:47.24 |-- /bin/sh /var/db/rrd/updaterrd.sh root 68133 0.0 0.0 14644 3244 - IN Thu17 3:33.10 |-- /bin/sh /var/db/rrd/updaterrd.sh root 70061 0.0 0.0 14644 3256 - IN 12Jun26 10:49.75 |-- /bin/sh /var/db/rrd/updaterrd.sh root 72652 0.0 0.0 14644 3256 - IN 25May26 20:48.76 |-- /bin/sh /var/db/rrd/updaterrd.sh root 77769 0.0 0.0 14644 3244 - IN Thu16 3:33.66 |-- /bin/sh /var/db/rrd/updaterrd.sh root 85039 0.0 0.0 14644 3252 - IN 23:27 0:00.37 |-- /bin/sh /var/db/rrd/updaterrd.sh root 86787 0.0 0.0 14644 3248 - IN Fri17 3:00.30 |-- /bin/sh /var/db/rrd/updaterrd.sh root 91678 0.0 0.0 14644 3248 - SN 15Jun26 9:04.32 |-- /bin/sh /var/db/rrd/updaterrd.sh
On systems with a large number of states (>100K) this can result in a significant CPU cycles being used to service the scripts.
Updated by Kris Phillips 16 days ago
- Status changed from New to Confirmed
I can confirm this behavior. If the previous updaterrd script fails to finish, subsequent ones will still spawn anyway. This leads into an infinite feedback loop of CPU utilization until the firewall crashes or becomes unresponsive.
Updated by Christian McDonald 16 days ago
- Status changed from Confirmed to In Progress
- Assignee set to Christian McDonald
We're working on a fix for this for 26.07.
Updated by Christian McDonald 13 days ago
- Status changed from In Progress to Feedback
Updated by Jim Pingle 7 days ago
- Subject changed from Multiple updaterrd.sh processes can be started to Multiple ``updaterrd.sh`` processes
Updated by
Actions