18 Replies Latest reply on Dec 10, 2009 1:15 PM by sedmo

    A whole host of issues I have with NPM

      I have only been using this product for 2 days in a test environment and I'm getting less and less impressed as I go. The sales engineers keep telling me how easy everything is and how you can customize "anything", but every time I ask them how I change a setting here or there they tell me "you can't"

      My main and biggest beef is the complete inability to add SPECIFIC graphs, volumes, or interface into the Network Summary main page. Or any type of "main" page for that matter. Sure, I can choose to create an "interface" or "volume" page, but it must be one of the other, not both? Why?

      Plus, when I create those above mentioned pages I seem to only be able to add graphs of interfaces or volumes from a single node, not multiple nodes of my choosing. Why?

      Why can't modules span multiple colums?

      Why do the volume graphs display a 95th percentile line? What is the point of 95th percentile on volumes?

      Why can't I turn on or off the 95th percentile bar?

      Why can't I change the color of graphs such as "average disk space used". Who decided I wanted my graph to be purple?

      Why can't I change the graph type of the same "average disk space used" graph. Maybe I don't want a bar graph, maybe I want an area graph.

       

      Ideally all those above issues I have I couldn't care less about IF someone can tell me how to make my main summary page display the following items.

      average bits per second - last 7 days : from my pix outside interface

      maximum traffic today : also from my pix outside interface

      network wide total bytes transferred - last 7 days : I want a table, not a graph

      average disk space used - last 7 days : Volume X from Node Y

      Available space on each volume table (found in the reports section) for ONLY Nodes X, Y, and Z

        • Re: A whole host of issues I have with NPM
          Donald_Francis

          Just from a longtime user perspective it is very customizable.  But it is done through layers.  Everything is built off of views (every page) and all views have resources on them.  The resources can be filtered in almost anyway.

          For example I made a view for my night operators that shows JUST my 5 largest sites.  This was done by putting the "all nodes" resource on my view and filtering that resource by name with the 5 sites.

          I would suggest hitting up your sales rep to assist you getting the info you need or finding the people to get it to you.

          • Re: A whole host of issues I have with NPM

            The way that I get around this is to create a new summary page that has all that you want on it and assign that as the home page in the account manager for which ever account you want to see that.

            • Re: A whole host of issues I have with NPM
              byrona

              hkor

              Your post heading indicates that you are having a lot of problems with the Orion software but after reading your post, most of your frustrations seem to center around the graphs, is this correct?  Are there other problems that you are also experiencing that you have not mentioned?

              I would also be curious to know which product(s) you are coming from so I have an idea of what kind of features and capabilities you are used to having available.

              We just recently moved to Orion and in the product selection process we reviewed about a dozen products and what we came to realize is that every product has it's warts, you just need to find a product that best fits your needs.

              I (and I am sure many others here in Thwack) would like to be able to help you with Orion in any way that we can.  If you could maybe let us know what your environment looks like and what you were hoping to do with Orion we will likely be able to help.

              • Re: A whole host of issues I have with NPM
                bcjordo

                So, I have a few suggestions.  I'm not sure if they are specifically what you are looking for but they have solved similar issues for my installation.

                1) Including a specific node/volume/interface graph on a summary page. 

                I do this by adding a custom html resource on the summary page and then inserting an iframe referencing the specific graph I'm looking for.  To get exactly what you want, go to the chart that most resmbles what you want and customize the timeframe/frequency/size etc and then view the printable version.  Use the URL it sends you to as the iframe source and you should be set.  (I can go into more detail if necessary)  You should be able to get several of your items this way

                2) Network wide total bytes transfered

                Use the same procedure above but this time after customizing your graph (total bytes transfered) for your timeframe and sampling interval, you can select "chart data" at the bottom of the customization screen.  This will output the graphical information as text.  You can include this as an iframe on a summary page as well.

                3) Reports as a resource

                Any report can be used a resource on summary pages.  There is a resource that allows you to do this.  You just have to modify the report you want (or make a copy if you don't want to modify the included report) and then include it as a resource.

                Hopefully this helps.  Let me know how it goes.

                • Re: A whole host of issues I have with NPM

                  I have been asked what environment I am coming from. Cacti is the polling system I was using before.

                  After another entire day of messing with Orion I can say I am still less impressed with it's "customizations"

                  I was able to display my pix outside interface graph per the suggestion of making it into an iframe however I find it a very cumbersome and ugly workaround. NPM should support this kind of functionality naively.

                  I was also able to get the average disk space report onto my network summary page via creating an entire new report in the report editor program. While I am amazed at the level of control the report editor has I am turned off by the fact pretty much 99% of orions customizations have to be down outside of the web interface. Whats the point of that?

                  I'd also like to point out that all my terabyte volumes are being reported incorrectly inside orion. I have a 5.45TB RAID50 volume being monitored and orion keeps reporting the total size as 5.999TB with 740GB free. However the operating system of the node reports that the actual size is 5.45TB with only 638GB free. Any help on that one?

                    • Re: A whole host of issues I have with NPM
                      byrona

                      hkor

                      First to address the incorrect volume size problem.  I am guessing that this device uses NetSNMP and if that is the case this is a reported bug with NetSNMP reporting the information incorrectly, Orion is just getting it's information from there.  I am well aware of this problem because the people that make OpenNMS opened the bug report with NetSNMP because of a problem I was having with OpenNMS doing the same thing.  If you are not using NetSNMP then I don't know what the problem could be here.

                      As far as graphing data is concerned, you have the UnDP (Universal Device Poller) that you can use to setup custom SNMP data collections and then present that data in Orion as well are run reports and set thresholds against it.  Have you had an opportunity to play with that yet?

                      I am familiar with Cacti, it uses RRD and that kind of graphing functionality is a feature request that has been submitted to SolarWinds via the following thread...

                      Multiple pollers - one graph

                      Hope some of this helps!

                      • Re: A whole host of issues I have with NPM
                        byrona

                        It sounds to me that you may be in the process of evaluating NMS systems, if this is the case the best advice I can give you (after having just done the same thing) is evaluate several solutions from different providers so that you can see how they all compare and select the one that works best for you.

                        We have found Orion to have a very large feature set at a very resonable price, the APM (Application Performance Monitor) is a great module for using your own scrips to monitor stuff as well as tons of Out Of The Box components for monitoring all sorts of different applications.  We actually plan to use the APM to get around the issue with NetSNMP reporting disk space incorrectly on large volume sizes.

                          • Re: A whole host of issues I have with NPM

                            I'm not using NetSNMP. I am using the default SNMP agent installed with Windows Server 2003 R2.

                              • Re: A whole host of issues I have with NPM
                                byrona


                                I'm not using NetSNMP. I am using the default SNMP agent installed with Windows Server 2003 R2.

                                 



                                That is interesting, sounds like the Windows SNMP Agent may have a similar problem with large volumes.  We don't have any Windows systems with drives that big so I wouldn't have seen this problem.  It would be interesting to grab that data with a different application, my guess is that it would still be inaccurate.  All Orion does is grab the SNMP data and store it in the DB so the numbers you are seeing are the numbers that Orion received from the SNMP agent (at least that is how it was explained to me and my experiences have supported that).

                          • Re: A whole host of issues I have with NPM

                            I know that can't be true because I have been using cacti for over a year and it has always reported the 5.45TB volumes as exactly 5.45TB.

                            I suggest getting some volumes in your dev lab that are 3, 4, 5TB and test this yourself.

                             

                            EDIT: Actually it doesn't even look like it's a large volume problem. It looks like all drives are being reported incorrectly.

                            Each VM has a 20GB volume, but orion reports it as 21.46. A vm with 17.95 used reports 2.05GB free, however Orion reports that same VM as having 2.211GB free.

                            I can't recommend this product if I can't trust its results.

                              • Re: A whole host of issues I have with NPM
                                byrona

                                hkor

                                Which virtualization solution are you using?  VMWare, Hyper-V, or something else?

                                • Re: A whole host of issues I have with NPM
                                  byrona

                                  Also, it's possible that your Cacti system is looking at a different OID or set of OID's than Orion is.  I know that with SNMP there are several different ways to get Hard Drive space and each one will give slightly different results.

                                  As I said before, I am pretty sure Orion is not doing any manipulation of the data, just storing what it receives from the SNMP Agent, maybe one of the SW folks can comment on the accuracy of this statement.  =)

                                  Myself coming from an NMS where I was able to specify which OID's and what mathematical functions were to be performed on the data (much like you can with Cacti) I can say that my data was slightly different as well.

                                    • Re: A whole host of issues I have with NPM

                                      I am using server 2003 VM's hosted on ESX 3.5

                                      (which by the way Orion reports my ESX info wrong too)

                                       

                                      After using this product more and more here are even more "hosts of issues" I have with this product.

                                      Custom Reports with Custom Graphs doesn't work
                                      Custom Reports with Custom Graphs

                                      Modules can't span multiple colums

                                      Drive space being reported incorrectly

                                      Difficult to change colors of graphs

                                      Poor out of the box APC PDU support

                                      Combine graphs into single view (2 interfaces from different nodes summed together)

                                      Multiple graphs from multiple nodes cannot be displayed on same page natively

                                      Difficulty setting up SQL filters in web interface to include NOT statements
                                      Cant use wildcards when using =, <, >, or !=

                                      Inability to customize "vendor" name sorting group

                                      Status overview of all nodes displays colum only instead of filling screen wasting screen real estate

                                      Inability to remove wasted screen space from graphs including "solarwinds" branding on each graph

                                      Does not pull correct CPU or Memory usage from VMware ESX 3.5

                                      Inability to set hard limits in line graphs (specify maximum graph height)

                                      Radial graphs waste too much screen real estate

                                      Inability to turn off 95th percentile graph lines

                                      Node Management page should have an "expand all" "collapse all" for nodes

                                  • Re: A whole host of issues I have with NPM
                                    Andy McBride

                                    Khor,

                                    We would like to help you but this is beginning to look like a spam post. This is the 4th time you have posted the exact same thing even though many in our user community have reached out to help. Please respond with more specifics and open a support ticket.

                                    Andy

                                    • Re: A whole host of issues I have with NPM

                                      A spam host huh? I'll be sure and let Jeff Hurst know that his company forum seems to think users having problems is spam.

                                      All he kept talking about is how great the community is. Reached out? The only solid answers I have gotten to any of my issues are to use an Iframe to load the interfaces that I want on the network summary page.

                                      By the way, I figured out a better way to do it anyway instead of an iframe.

                                      I STILL have incorrect graphs fro volumes and ESX servers. Like I keep saying, cacti reports all volumes as the correct size with the correct free space available, but solarwinds can't seem to figure it out. Why should I buy this product again?

                                      • Re: A whole host of issues I have with NPM

                                        The answer to each of your questions about why the software does not work the way you would prefer it to is because it's not currently designed and built that way.  If a product does not meet the criteria outlined in your evaluation process then you should probably select a different product.