Yeah you are hijacking a thread, but whatever. I'll change the subject line and start a new thread :)
I have rancid and websvn running on the same box. That box is separate from observium. I did this because I wanted rancid to run on a vm. Having it on a VM gives me somewhat of redundancy. Depending on the number of threads and number of devices you give rancid, it could interfere with observium. The server that runs observium comes into play as well. The more devices and more ports you have, the more threads you need. Observium is much more important to me than rancid; however, rancid is really important. That is another reason why I separated them.
Then I have an NFS mount between rancid and observium so it can see rancid. I don't have a history of configurations in observium. I am not sure how that works. That's where websvn excels. With websvn the network guys can see what changed between revisions. So with websvn, I didn't spend much time on setting up history of configuration changes in observium.
I haven't setup smokeping. The network guys always tell me they don't care about icmp testing to determine latency. Mainly because some routers/switches, that participate in QoS, will make icmp lowest priority and put the router traffic, voice traffic, and sometimes video traffic (depending on the device) higher priority. The part that I love though is when I am working with them to troubleshoot a network problem, they want a trace route. I always say to them well I thought you didn't care about icmp and trace route uses icmp. :) Inside observium in the graph section of the device in the subsection of poller, it shows icmp response. I think that is enough for me and the network guys.
Mike
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Peter Casserly Sent: Friday, June 13, 2014 8:03 AM To: Observium Network Observation System Subject: Re: [Observium] Rancid can't get CVS working
For those using a WebSVN or similar application: Are you running that on a separate server or all on the same server? I have Observium and rancid running fine but want to add SmokPing and WebSVN.
I don't want to create a mess running too many services running on one server.
Looking for feedback on best practices/success vs. failure stories. Apologies in advance if this reply is off topic or hijacking the thread.
Mike,
That is helpful. Thank you for your time and response.
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Teixeira, Mike Sent: Friday, June 13, 2014 11:35 AM To: 'Observium Network Observation System' Subject: [Observium] Smokeping, Rancid and Observium
Yeah you are hijacking a thread, but whatever. I'll change the subject line and start a new thread :)
I have rancid and websvn running on the same box. That box is separate from observium. I did this because I wanted rancid to run on a vm. Having it on a VM gives me somewhat of redundancy. Depending on the number of threads and number of devices you give rancid, it could interfere with observium. The server that runs observium comes into play as well. The more devices and more ports you have, the more threads you need. Observium is much more important to me than rancid; however, rancid is really important. That is another reason why I separated them.
Then I have an NFS mount between rancid and observium so it can see rancid. I don't have a history of configurations in observium. I am not sure how that works. That's where websvn excels. With websvn the network guys can see what changed between revisions. So with websvn, I didn't spend much time on setting up history of configuration changes in observium.
I haven't setup smokeping. The network guys always tell me they don't care about icmp testing to determine latency. Mainly because some routers/switches, that participate in QoS, will make icmp lowest priority and put the router traffic, voice traffic, and sometimes video traffic (depending on the device) higher priority. The part that I love though is when I am working with them to troubleshoot a network problem, they want a trace route. I always say to them well I thought you didn't care about icmp and trace route uses icmp. :) Inside observium in the graph section of the device in the subsection of poller, it shows icmp response. I think that is enough for me and the network guys.
Mike
From: observium [mailto:observium-bounces@observium.org] On Behalf Of Peter Casserly Sent: Friday, June 13, 2014 8:03 AM To: Observium Network Observation System Subject: Re: [Observium] Rancid can't get CVS working
For those using a WebSVN or similar application: Are you running that on a separate server or all on the same server? I have Observium and rancid running fine but want to add SmokPing and WebSVN.
I don't want to create a mess running too many services running on one server.
Looking for feedback on best practices/success vs. failure stories. Apologies in advance if this reply is off topic or hijacking the thread.
participants (2)
-
Peter Casserly -
Teixeira, Mike