warbird

Now, the point is if we can do something else until we can get the phisycal server assigned? Reduce your statistical polling interval. If you are running NTA, reduce the number of flows pointed at your primary poller. I mistyped. To clarify what I meant: You can decrease load on the server by slowing down how often the…

Yup. Our issues were due to too much load on one or both of the polling engines. I was unable to determine if it was the primary polling engine or the secondary but as soon as we brought a 3rd poller online and I reduced the load on the existing pollers, the strange issues we were seeing went away. The curious thing about…

Now, the point is if we can do something else until we can get the phisycal server assigned? Reduce your statistical polling interval. If you are running NTA, reduce the number of flows pointed at your primary poller.

My file size was ~37MB. After taking the noted steps and giving it about an hour, the file is back up to ~13MB. Didn't notice any huge changes in web site performance but everything is working, which is good. The BusinessLayerHost.exe service was at ~1.2GB memory consumed this morning before I made the change. It is…

Well, I gave it a shot anyway. End result: absolutely no change. Ah well. If anyone knows of other ways to reduce the memory footprint of w3wp.exe, post it here?

Thanks Mike. Yes, I am running 9.5 SP4. I don't think what I am seeing is the exact problem you were seeing, only related. I just want to reduce the memory footprint of w3wp.exe and came up with this method. Wondering if anyone else has tried it and/or what opinions are on it.

What does being a beta tester for IPv6 entail? Do we have to run a new rev of NPM or is it more like a module/plugin?

Hmmm. OK. They had already closed my ticket. I will reply back, asking them to reopen it and give me the details. Any reason you can't post them here?

Thanks for your help, as well, David. Things are still running very smoothly this morning. It should be noted that the AWE setting will only make a difference in 32 bit Windows and SQL.

Just wanted to say thank you to njoylife and provide a quick update... implementing AWE appears to have completely addressed the disk queue length problem. We are running 32 bit Windows Server Enterprise but all processes still competed for only 2 GB of memory. I implemented AWE, leaving the min/max settings at default so…

In my case, 3 things lead to this problem. 2 of which I was able to address, 1 of which is not addressable. The first 2 were load issues and db issues. My polling engines were being taxed beyond their limits and I had very large tables with very old data in my db that were not rolling up properly (we are currently…

Also njoylife, How did you partition up your memory with AWE? Did you devote about half to it and leave the rest dynamic? Thanks again, Warbird

Dave, This is was the suggested solution I got. I just implemented it this morning. It will take a few days to see if it has resolved the 2 issues I've been seeing. Will post again with an update late this week or early next week and verify the solution if it is working: *** Please navigate to Documents and Settings\All…

Jeff, do you have a very high number of people hitting your web front end?

I'm running netflow, orion, and NCM on the same DB, as well as some other app's DBs. it is 4 proc/8 core, 32 GB RAM (only using 8-10) and cpu usage is pretty low. I am connected to SAN, and we did run into problems a while ago, but it was resolved by simply configuring the SQL DB memory to utilize AWE. Test this and you…

np. I have also had to evaluate the importance of monitoring every singe interface on every single node. A nice common ground may be to monitor every node and only monitor interfaces that are trunk ports or very important server ports. Good luck!

I have also noticed this problem. Because my polling engines are running on 32 bit Windows, even though I have 6GB of RAM in the system, Windows only knows how to utilize 2 GB total. After upgrading to 9.5, I have seen the w3wp.exe process consuming ~300MB, which maxes out the 2 GB the system can use. I then notice…

Don, sorry for the delay in responding. Yes, I did get this issue resolved. It involved working closely with support and cleaning up some things in my db. I do not remember all of the details. I recommend you open a ticket with support. They should be able to reference the above ticket for the solution.

That is actually .02 and that is a very good number to see. If it were "50", it would actually read something like 50.04 or whatever. When I first inherited our Orion implementation, I was seeing numbers between 400 - 800. No joke. Right now, I am looking at a single spike within the measurement window such that the…

It appears this issue was being caused by a couple of tables in the main NetPerfMon db that had issues. I believe this stemmed from the old wireless module, from an older version of NPM. The tables in question were the largest tables in my db and had very old data contained within. We truncated those tables, allowed db…

Nope. I avoid VM's like the plague when it comes to something as system intensive as Orion. How many other VM's are running on the physical server? 

Ah, yes. The SAS drives in your SQL server are what is making it all work well for you, IMO (and you are likely utilizing a good deal of RAM). We are running 7200 RPM SATA drives and after all the research I've done, I am surprised they are performing as well as they are, even in the RAID 10. I can't wait to get 6 of the…

No problem. A nice feature would be to disallow addition of nodes that are not polled by the same poller that IP SLA is installed on. Not certain if that is possible but it would be nice!

Renaming the operation, as you described, works great. The only complaint is that I have to manually do this for a couple hundred different operations. Thanks for helping with this. Ended up being trivial but my SLA view sure looks a lot better for it!

aliendan: The disks are 7200 RPM, 750 GB SATA disks. There are 6 disks total in the array and the OS 'lives' on the same array as the SQL db. The controller is a PERC 6i w/256MB cache. I currently have the 'read policy' set to "Adaptive Read Ahead" and the write policy set to "Write Back". Here's the kicker... running…

What other VM's will be running on the physical server? Also, do you anticipate any drastic growth of your infrastructure? I toyed with the idea of moving to VM's but given our number of elements and future growth considerations, it simply isn't practical for us.

David, Are they attributing the ongoing slowness issues you are still seeing to the hardware problems that messed up your upgrade, or is it something else? I've been running 10.0.0 SP1 for a few weeks and have been happy. Was considering the upgrade to 10.1 but have decided to hold off for now, pending results from you and…

Thanks Andrew! Brandon, our time line is this spring. It is looking like around May is when we will be turning it up.

BTW: I just saw you said you moved your db to a SAN. Is your SQL server connected to it via fiber? How many spindles? Is your SQL server 32 bit Windows? Just moving my db to a very high end SAN wasn't good enough. I also had to implement AWE to allot SQL more memory.

To get an idea of what your avg disk queue length is doing, during a busy part of the day for your network, log into your SQL server, click start, run, type perfmon, and hit enter. One of the default counters already in place is the Avg Disk Queue Length. Click the little light bulb button to highlight the counter you have…