Avert your eyes people, no fix for this one in the post… (yet) I just wanted to inform the community about a little bug / error with the syslog service on vSphere / ESXi 7 which I stumbled on.
I started a little thread over at the VMware community forums, between this blog post and that thread hopefully VMware will listen.
I tested this on vCenter 7.0.1 16858589 & ESXi 7.0.0 15843807
Error: A specified parameter was not correct: id
How to replicate? in the vCenter or ESXi GUI just try to stop the syslog service on a host. The GUI’s will generate an error, see below.
SSH onto the host and check the syslog service by running the command below and the following error is generated.
Watchdog fork failed 28 (No space left on device)
The only way to stop the service is by killing it, not optimal but works.
ps -Cc|grep vmsyslogd
17375454 17375454 grep grep vmsyslogd
17275879 17275879 vmsyslogd /bin/python /usr/lib/vmware/vmsyslog/bin/vmsyslogd.pyc start
17275878 17275878 wdog-17275879 /bin/python /usr/lib/vmware/vmsyslog/bin/vmsyslogd.pyc start
kill -9 17275878
kill -9 17275879
Thanks to @virtuallywired –> virtuallywired.io for testing it an confirming it on 7.0 and also on Update 1.
hope you found this helpful.
Hey vMAN, ESXi 6.7U3 Also exhibits this issue, I’ve updated the community thread with a screenshot.
FYI I wrote a big post here to help people and failed the crappy captcha and your blog dumped my entire post.
TL;DR version: if you are getting an implementation error try changing your “local log output” for syslog to a directory that actually exists. After changing it to /tmp I was able to configure the “remote host”
Sorry about the captcha, thanks for the info