doesn't seem to be running. I put an "exit" in each init scripts (elasticsearch, logstash, logstash-web) so they don't do anything but logstash still runs after a reboot
Arthur40A joined the channel
seems to be started from elsewhere
and dies but respawns
nat2610 has quit
imperfect-
on debian systems is it /etc/init.d ?
or rc.d
look at hte scripts of your runlevel
fdrolet
init.d
imperfect-
grep them for java
loggyer has quit
fdrolet
only those 3 I mentioned
imperfect-
well
remove all threee scripts from the runlevel directory
reboot
plasmonyc has quit
fdrolet
just did that, even shut the vm completely. still there 0_o
MartinCleaver has quit
mschmitt has quit
volter joined the channel
olivier__ joined the channel
walterheck joined the channel
walterheck has quit
found the culprits. there where executable scripts logstash and logstash-web in /etc/init
walterheck joined the channel
smerrill is now known as smerrill-offline
rbergeron joined the channel
in addition to /etc/init.d
first time I see that
zbp has quit
jalaziz joined the channel
shubhang has quit
they are called logstash.conf and logstash-web.conf, so I'm not sure if they are supposed to be executable
srage joined the channel
shubhang joined the channel
mschmitt joined the channel
srage_ has quit
kanarip has quit
kanarip joined the channel
jerryitt has quit
nat2610 joined the channel
OK so after moving away /etc/init/logstash.conf and logstash-web.conf, logstash doesnt run after a reboot, so that's one mystery solved (for me anyway). Now I have to understand why the processes keeps respawning and using lots of cpu time
mleung has quit
must be something wrong with my config, but as I followed digitalocean's guide and don't see any error in the logs, I don't really know where to go. Thanks for your help on that first part anyway :)
newmember has quit
i really want to use this elk stack because it seems awesome. I did some filters, setup some forwarders and it's great. It still works, I see things in Kibana, but I know this high cpu usage isn't right. Don't want it to bite me after I take the time to setup filters and forwarders for all kinds of services and servers
Prune has quit
sorry I'm probably rambling
kanarip has quit
incry6t1 joined the channel
kanarip joined the channel
mdelnegro has quit
mdelnegro joined the channel
ph has quit
ph joined the channel
torrancew
fdrolet: just catching up (was at lunch) but it seems you're an ubuntu user. Upstart is the init system Ubuntu uses, and it looks at /etc/init/*.conf
logstash ships /etc/init.d/ scripts for things as well, and this is where it all goes wrong for ubuntu folks
nat2610 has quit
Typically, ubuntu uses a wrapper script as a symlink in /etc/init.d, so that "service foo stop", or /etc/init.d/foo stop still cue upstart
but logstash is shipping a real init script there, since it uses that on non-upstart platforms
fdrolet
torrancew: you are right, this is on ubuntu, but I had the same problem on debian
torrancew
fdrolet: which debian version?
fdrolet
7.7
torrancew
I have not seen that behavior, and am a deb7 user
Bliked
Any advices for list vs channel with Logstash / Redis?
torrancew
if you can share your findings in real-time, happy to look into it with you
ph has quit
fdrolet
torrancew: great, thanks for these init pointers, I am still confused with it sometimes, even after 10+ years of linux admin ;)
tchiang joined the channel
stasher joined the channel
torrancew
:) no worries
bozz has quit
fdrolet
so are you saying that there are 2 different init scripts (one in /etc/init and one in /etc/init.d) and that breaks ubuntu upstart?