Congo

SP1 is now available. Denny, When can we expect this to be on the customer portal? I have some very serious support tickets that were put on hold to await this service pack, so I am amped to try it out.

 Wanted to express my interest on TestTrack # 8047 also. I have 5 custom pollers that I am waiting to alert on table values.

 I would also like to express my interest in getting advanced alerting capability in syslog and trap viewer; plus the ability to parse information from traps. I also am currently having to use an open source trap receiver that has loads more functionality than trap viewer, when this should be included into for-profit…

 No compression, and 64b counters are enabled. I wasn't referring to bursting situations either, these are core router links, edge atm links, it does not seem to be prejudice.

I have been looking for this exact functionality since we started with Orion 18 months ago. I have yet to find a valid solution out of support, or even two different sales engineers. We need traps just like any one else, and need them (and syslog actually) to be tied into advanced alerts. Our NOC, customers, and management…

 I do the same monitoring, and have been unable to figure out the email syntax also. For the time being, I just have it give me a router alias and that a BGP state has changed. From there I just go take a look at the router.

3 years? Borrow the CERN's cloud computing.

I was curious of the same thing. That is something like 30,000 each poller! We have a poller that monitors 4 nodes, but 10,000+ interfaces...and it can only poll every 30 minutes or gaps appear.

What I object to is the removal of existing functionality for no apparent reason. What does it hurt to keep the system manager as it is until the web console is up to par - even if you never add new functionality to the system manager? I agree with this. Keep the functionality of NM where it is at, and keep bringing the…

I find this issue all the time also, actually. MRTG will display what we think is correct, and our FlukeNetwork's Netflow Analyzer and Orion will display different values entirely from each other, both over utilizing the circuit.

Well, I am not quite sure that processor is the bottle neck for elements. I think there are a lot of network conditions that come into play for that. The Netflow solution we use is Netflow Trakker from Fluke Networks. The granularity difference is necessary for our operations, as we upstream provide up to 2.2gbps. Not sure…

We are working on an SP now which addresses the issues with the variables and also fully documents them as well. ETA? I see a knowlege article in this thread that mentions SP3 already. Is it days away?

Ouch jeff.stewert My database is the other way around, that is what my Write performance looks like(average 169), while my read performance averages .02 I am still attempting to convince my management that solid state is necessary.

I am quite interested in this new UI you are working on Denny. Currently, in a fast changing environment with our 18,000 elements, it is almost a full time job to keep changing around alerting to keep things properly suppressed. I have all but given up.

I was involved much in the discussion of this a long time ago like ecornwell mentions here. Since this discussion forever ago, I have adapted somewhat to the usage of the web console, and aside from the speed issues (which we are attempting to resolve ourselves) I find it less heinous than before. One thing to add that I…

In total we have ~18000 elements, plus APM. Box1 = Website, APM(800+), Voip(10), Wireless(10), ~3500 Elements Box2= ~5000 elements Box3= ~9800 elements Box4= SQL The above works quite well for us. Box1 and Box3 are Dual quadcore 16gb RAM...which is overkill. Box2 is a virtual machine in a cluster of about 6 machines with…

I was in the process of opening a ticket, and went back to attempt this again and screenshot the error...however it worked as intended! I did updates to SP4 yesterday, so not sure if that did it...but this is no longer an issue!

Yes - I'm looking for changed to exsisting, change controlled, systems. the changes I see in Solarwinds would be compared against what was documented and approved. Unapproved changes could then be investigated. Nail on the head here. I want both this, and then a way to track new devices, and investigate them as well. We…

 Your problem is the database is located on the same machine as the poller and web....and only 2gb of ram. Put the DB on another machine, and you will see performance spike. Double the ram and watch it only get better.

As part of this, let's say we have 30 switches at a facility prior to the revamp. Now that the changes are all implemented, I need to go into Orion and delete 5 switches, delete 10 interfaces on other switches, add another 10 switches, add new interfaces on 10 of the remaining switches, and add 10 wireless AP's. OK,…

Our most taxxed poller is at 10160 elements, spread across 3 nodes. (so 10157 interfaces) We actually have this on physical hardware: Intel 5130 Quadcore / 8GB Ram This connects back to a database with the other 9000 elements, and performance here is good. This poller sits at about 8% CPU and 15% Mem all day long. Our…

All of the results. So, jitter, MOS, packet loss, RTT, etc.

I was really hoping to see the alert dependencies rework in this release, for help with alert suppression. Last I heard, there was other work to be done first structurally to allow this. I know that ALL of your customers can benefit from this, not just a select few; I figured that would weight this differently.

Im shocked also. One spindle for that tight of polling and that many elements? IS this one drive a Fusion IO ? : )

We are at 18000+ elements (and around 800APM) across 3 pollers. I LOVED 9.5 , and our performance is about the same. Some things are faster, while other times there are random spikes of waiting a minute to load the home page. Granted, I had to open 8 or 9 tickets soon after the upgrade...that was a bit expected.

Bumping this back up. Any ideas Brandon? Or has anyone else had this issue?

I should probably clarify. 5760x1920 is one particular view that I use, between laptop and additional monitors. We also display on regular LCD's 1280x1024 in the NOC, and large monitors at 1920x1080.

We use Orion to monitor our own network, and then we offer to delve past the WAN that we provide to our customers, and monitor their LAN. We do this over VPN's, and back to our Orion cluster. We have found this works better than remote polling deployments. I would rather have poll data going across the wires instead of…

 Good to know I am not the only one cycling tabs out there, Grant.

I do something similar, and I use Netflow to achieve this. However, I use a program called Netflow Trakker from Fluke Networks. Much higher granularity than the Solarwinds equivalent. In your situation, I think that you would want to use some clever alerting on % utilization or even traffic rate and have it alert you when…