Avamar: Checkpoint failed with result "MSG_ERR_BADTIMESYNC."

Checkpoint failed with result "MSG_ERR_BADTIMESYNC"

The avmaint cpstatus command shows the following error:

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<cpstatus
  generation-time="1663935384"
  tag="cp.20220923121551"
  status="error"
  stripes-completed="0"
  stripes-total="0"
  start-time="1663935351"
  end-time="1663935351"
  result="MSG_ERR_BADTIMESYNC"
  refcount="1"/>

The mapall --parallel 'date' command shows that a node is out of sync (To run the mapall command, the keys must be loaded per article Avamar: How to Log in to an Avamar Server and Load Various Keys)

admin@utility:~/>: mapall --parallel 'date'

Using /usr/local/avamar/var/probe.xml
(0.0) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.2 'date'
(0.1) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.3 'date'
(0.2) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.4 'date'
(0.3) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.5 'date'
(0.4) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.6 'date'
(0.7) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.9 'date'
(0.6) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.8 'date'
(0.5) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.7 'date'
Fri Sep 23 13:05:21 UTC 2022
Fri Sep 23 13:05:21 UTC 2022
Fri Sep 23 13:07:17 UTC 2022 <---- out of sync node
Fri Sep 23 13:05:20 UTC 2022
Fri Sep 23 13:05:22 UTC 2022
Fri Sep 23 13:05:20 UTC 2022
Fri Sep 23 13:05:22 UTC 2022
Fri Sep 23 13:05:21 UTC 2022

Verifying the network time protocol (NTP) with the ntpq -pn command shows the message "Connection refused" on the suspect node (Output edited to show only the affected node):

admin@utility:~/>: mapall --noerror '/usr/sbin/ntpq -pn'

Using /usr/local/avamar/var/probe.xml
...
(0.3) ssh -q  -x  -o GSSAPIAuthentication=no admin@192.168.255.5 '/usr/sbin/ntpq -pn'
/usr/sbin/ntpq: read: Connection refused
...

When verifying the status directly as root on the affected node, the network time protocol daemon (NTPD) shows "Active: activating (auto-restart) (Result: resources)":

root@node03:~/>: systemctl status ntpd

● ntpd.service - NTP Server Daemon
   Loaded: loaded (/usr/lib/systemd/system/ntpd.service; enabled; vendor preset: disabled)
  Drop-In: /run/systemd/generator/ntpd.service.d
           └─50-insserv.conf-$time.conf
   Active: activating (auto-restart) (Result: resources) since Fri 2022-09-23 13:22:35 UTC; 1min 58s ago

The status should reflect Active: active (running):

● ntpd.service - NTP Server Daemon
   Loaded: loaded (/usr/lib/systemd/system/ntpd.service; enabled; vendor preset: disabled)
  Drop-In: /run/systemd/generator/ntpd.service.d
           └─50-insserv.conf-$time.conf
   Active: active (running) since Fri 2022-09-23 14:04:37 UTC; 26s ago

Attempting to start NTPD fails:

root@node03:~/#: systemctl start ntpd.service

Job for ntpd.service failed because a configured resource limit was exceeded. See "systemctl status ntpd.service" and "journalctl -xe" for details.

The output of the journalctl -xe command reports "No space left on the device" messages.

The df command shows that /var is at 100%:

df -kh

Filesystem      Size  Used Avail Use% Mounted on
devtmpfs         16G  8.0K   16G   1% /dev
tmpfs            16G     0   16G   0% /dev/shm
tmpfs            16G   50M   16G   1% /run
tmpfs            16G     0   16G   0% /sys/fs/cgroup
/dev/sda5       9.8G  2.4G  7.0G  26% /
/dev/sdg1       183G  8.3G  165G   5% /ssd01
/dev/sda1       979M   50M  878M   6% /boot
/dev/sdd1       1.9T  236G  1.6T  13% /data04
/dev/sdc1       1.9T  240G  1.6T  13% /data03
/dev/sde1       1.9T  236G  1.6T  13% /data05
/dev/sdf1       1.9T  238G  1.6T  13% /data06
/dev/sdb1       1.9T  238G  1.6T  13% /data02
/dev/sda7       2.0G  2.0G     0 100% /var       <------- 100% Use
/dev/sda3       1.8T  267G  1.6T  15% /data01

NTPD relies on /var/lib/ntp/drift/ntp.drift which contains the latest estimate of clock frequency error.

If /var is at 100% full, NTPD cannot update or create the ntp.drift file, and NTP does not function correctly.

1. On the affected node, investigate and resolve 100% usage of /var.
2. Once corrected:
   1. Restart NTPD: 

root@node03:~/#: systemctl restart ntpd

1. 2. Check the status of NTPD:

root@node03:~/#: systemctl status ntpd

Results similar to the following should be seen:

● ntpd.service - NTP Server Daemon
   Loaded: loaded (/usr/lib/systemd/system/ntpd.service; enabled; vendor preset: disabled)
   Active: active (running) since Tue 2022-09-27 21:21:42 UTC; 37s ago
     Docs: man:ntpd(1)
  Process: 29442 ExecStart=/usr/sbin/start-ntpd start (code=exited, status=0/SUCCESS)
 Main PID: 29463 (ntpd)
    Tasks: 2
   CGroup: /system.slice/ntpd.service
           ├─29463 /usr/sbin/ntpd -p /var/run/ntp/ntpd.pid -g -u ntp:ntp -c /etc/ntp.conf
           └─29464 ntpd: asynchronous dns resolver

Sep 27 21:21:42 node03 ntpd[29463]: Listen normally on 3 bond0 10.n.n.52:123
Sep 27 21:21:42 node03 ntpd[29463]: Listen normally on 4 bond1 192.168.255.22:123
Sep 27 21:21:42 node03 ntpd[29463]: Listen normally on 5 lo [::1]:123
Sep 27 21:21:42 node03 ntpd[29463]: Listen normally on 6 bond0 [fe80::260:16ff:feaa:2a10%11]:123
Sep 27 21:21:42 node03 ntpd[29463]: Listen normally on 7 bond1 [fe80::260:16ff:fea9:b182%12]:123
Sep 27 21:21:42 node03 ntpd[29463]: Listening on routing socket on fd #24 for interface updates
Sep 27 21:21:42 node03 start-ntpd[29442]: Starting network time protocol daemon (NTPD)
Sep 27 21:21:42 node03 systemd[1]: Started NTP Server Daemon.

1. 3. Verify NTP with ntpq:

root@node03:~/#: /usr/sbin/ntpq -pn

Results similar to the following should be seen:

     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*10.241.216.209  10.233.131.242   2 u  966 1024  377    0.558    1.559   0.600
+192.168.255.21  10.241.216.209   3 u  401 1024  377    0.152    0.521   0.420

3. Confirm the resolution by running a manual checkpoint from the Avamar Utility Node.