You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I followed instructions in Installing PuppetDB from source for my ubi8 image and install.sh script was giving errors. I started investigating, and it showed that this part of script
(maybe_dead=''
pid="$(pgrep -f "puppetdb.* -m puppetlabs.puppetdb.main")"
if test $? -eq 0; then
kill "$pid" 2>/dev/null
for i in {1..75}; do # Wait up to ~15s
if kill -0 "$pid" 2>/dev/null; then
maybe_dead=1
break
fi
sleep 0.2
done
if test -z "$maybe_dead"; then
echo 'Unable to kill puppetdb server cleanly; sending KILL' 1>&2
kill -9 "$pid"
fi
/sbin/service puppetdb start > /dev/null 2>&1
fi)
is what cause errors.
As i see from the comments above this part
ppetdb changed its main namespace in 6.3 and the current rpm
upgrade strategy fails to restart the server correctly because
t ends up using the templated stop from the new package which
refers to the wrong namespace for the older pdb that's still
running. Kill the old pdb if found and start the new one.
These changes can be removed in PuppetDB version 8.
And this part of script can be easily deleted without any consequencies
Expected Behavior
after compiling puppetdb-*/pkg folder with lein with-profile ezbake ezbake stage ./install.sh file should work without errors
Steps to Reproduce
Steps to reproduce the behavior:
On RHEL8 or ubi8 try folllowing install from source guide
./install.sh will give something like
Environment
Version [tag 8.7.0]
Platform [ubi8]
Additional Context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered:
I followed instructions in Installing PuppetDB from source for my ubi8 image and install.sh script was giving errors. I started investigating, and it showed that this part of script
(maybe_dead=''
pid="$(pgrep -f "puppetdb.* -m puppetlabs.puppetdb.main")"
if test $? -eq 0; then
kill "$pid" 2>/dev/null
for i in {1..75}; do # Wait up to ~15s
if kill -0 "$pid" 2>/dev/null; then
maybe_dead=1
break
fi
sleep 0.2
done
if test -z "$maybe_dead"; then
echo 'Unable to kill puppetdb server cleanly; sending KILL' 1>&2
kill -9 "$pid"
fi
/sbin/service puppetdb start > /dev/null 2>&1
fi)
is what cause errors.
As i see from the comments above this part
ppetdb changed its main namespace in 6.3 and the current rpm
upgrade strategy fails to restart the server correctly because
t ends up using the templated stop from the new package which
refers to the wrong namespace for the older pdb that's still
running. Kill the old pdb if found and start the new one.
These changes can be removed in PuppetDB version 8.
And this part of script can be easily deleted without any consequencies
Expected Behavior
after compiling puppetdb-*/pkg folder with lein with-profile ezbake ezbake stage ./install.sh file should work without errors
Steps to Reproduce
Steps to reproduce the behavior:
Environment
Additional Context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered: