mesverrum · Observability Architect · ✭✭✭✭✭

Comments

  • There is no public documentation of the table relationships, and you are correct there are many table columns that are not explicitly identified as being keys that you will have to comb through. Successfully combining two databases is extremely difficult and in most cases it doesn't go well. Hopefully you have plenty of…
  • Do your actions involve an escalation? If there is not then they basically all happen at the same time, or within milliseconds. It doesn't wait for step 1 to proceed to step 2. If you want to let the ncm action complete before your email goes out you usually need to add in an escalation and let it wait maybe a minute or so…
  • It can be done using a feature called a universal device poller. Any time there is an SNMP based bit of information you want to collect that Solarwinds does not collect out of the box you can just tell it the OID you want to have it check. This video demonstrates creating them SolarWinds Orion Network Performance Monitor…
  • Given that SWQL doesn't support most forms of explicit casting you just have to develop within the constraints you have. Why does your 2018-03-08 have to be manipulated as a string? Is there not a way to use any of the currently defined 34 SWQL datetime functions to get the same effect? SWQL Functions · solarwinds/OrionSDK…
  • Mostly the same Select Distinct data.Name, sum(data.TotalPowerLoad) as TotalPowerLoad, data.location FROM ( SELECT distinct (substring(n.Caption,1,16)) AS NAME, cs.ComponentStatisticData AS TotalPowerLoad, n.location FROM orion.apm.CurrentStatistics cs JOIN orion.nodes n ON n.NodeID = cs.NodeID WHERE…
  • Solarwinds is capable of receiving netflow data from all versions of Cisco. The key question I think you have is if your cisco devices themselves support netflow and the best way to verify that is just to talk to your vendor. For the most part switches do not support flow regardless of the license they have applied because…
  • encapsulationdown​ Do these switches have any kind of UPS? From the switch itself there is no way to distinguish between hard power failures and the device just not being able to answer a ping due to network or config problems, but via a SNMP enabled UPS or some other kind of environmental monitor you can monitor incoming…
  • Groups get their status from the members of that group, not any children of those members. To cause the effect you are asking for I believe you would have to actually add the interfaces you are interested in to the group. Not too hard in a small environment but could be a pain to do with a lot of groups.
  • The published recommendation is to make sure the ape has under 250 ms of latency between it and the SQL db.
  • At this time there is not a built in integration with other Meraki devices besides the AP's. You can enable direct SNMP polling for the firewalls and switches but from the people I spoke with you might find the available OID's are pretty limited compared to what you would see from most other firewalls. SNMP Overview and…
  • You could steal a play from the solarwinds built in alerts for Top CPU and Memory processes. Step 1, trigger alert and the action is to launch your script to tracert (l33t Tracer-T video​ ) add an escalation, wait maybe a minute or two for the script to run, now the trick is that one of the final actions of your script is…
  • The no data available placeholder isn't due to the system being slow in any way, it means that resource is charting out some metric that your object hasn't had during the applicable time window or Orion hasn't been able to collect the data. I've not seen that particular resource fail to have data though. I'm not 100% where…
  • I run into this once in a while in client environments, usually bouncing the services or running a configuration wizard on the primary server forces it to catch up with the latest properties list. I thought they addressed it in recent versions because I haven't had the problem in about a year.
  • There is currently no streamlined automatic failover/ha for the NTA database. The application part of NTA has absolutely has to be installed on the same server as NPM, it can't be run standalone. You also need to set up an additional separate server for the NTA database, it does not use SQL so it cannot live in the same DB…
  • I ran into this behavior last week also, haven't had a chance to try and identify the cause.
  • Nice, I'm surprised I missed the announcement of this new resource in Orion. I've been playing with the tiles in EOC for a bit and the biggest hurdle I had run into there with them was that I can build one tile, and that's good, and I can build 10 tiles and that's a little tedious, but then my clients see the view I put…
  • Presumably because there is something preventing you from reaching the switch with SNMP. Can you putty from the orion server to that switch's IP? Can you telnet from the orion server to the switch on port 161?
  • So what's going on is that your SAM script is setting the status but it's not populating all the other tables that would normally be populated when polling a node. Can't you just report based on the availability of the SAM script and basically get the same results?
  • Typically this is due to running an old firmware or version of OMSA, see mention here http://en.community.dell.com/techcenter/systems-management/f/4494/t/19440792 https://arstechnica.com/civis/viewtopic.php?f=17&t=83704
  • When they show up like that in Atlas it means the node/object that they used to be got deleted from the database. Deleted objects are purged every night when the maintenance runs at 2:15 am so unless the change just happened you are pretty unlikely to be able to find out what it used to be. If you hover over the object it…
  • The trick to doing this is to not select a specific map on the details page, you should be using the selection that says ${name} and then just save the map you want to show up with exactly the same name as the group, Orion will display whatever map it finds with a correlating name if one exists.
  • If it were my environment I'd pull the history on that undp poller to see what values were in the database (assumes you have "keep historical" checked on that undp). Normally there are no entries at all on a missed poll so there is nothing for the alert engine to act on and I wouldn't expect a message. I also think it is a…
  • The answer i heard on this is that the new streamlined installer basically replaced to UA, run it and it will let you know your upgrade path. 
  • Can be hard to pin them down after the fact in larger environments, that's why I always try to include a "This message generate by ${N=Alerting;M=AlertName}" in the body somewhere In any case, what I would do is look through the alerthistory table in the database and try to correlate those messages to timestamps or the…
  • in cases where you are doing a small enough deployment that you don't want to stand up a separate SQL server I would suggest you at least split the db drives off. Something like a drive for Windows system stuff, one for the Orion and SQL applications, one for SQL DB, one for NTA DB. Something like 4 vcpu and 16 gb of ram…
  • The way the logic works on table pollers is that it evaluates each row of the poller table separately, so the rule you have there would alert any time 10.0.1.24 is in any table poller at all. It doesn't check the whole table for those two values as a set. You could probably get away with specifying the rowid that currently…
  • When you add a subnet there is a check box for "disable automatic scanning" and as long as you don't check that (or don't have it globally disabled in the ipam settings) it should attempt to scan right away. Individual subnets > Global scan settings >
  • Those datasets are pulling swql queries, swql is 95% the same as sql so almost any method you find online for filtering SQL time sets would apply similarly. You node availability query already has a last 1 day filter applied with the getdate()-1 part See the SWQL function list here SWQL Functions · solarwinds/OrionSDK Wiki…
  • Minimum required permissions for vCenter polling in NPM - SolarWinds Worldwide, LLC. Help and Support Read only Plus browse datastore
  • I work as a consultant so I don't get to own any of the environments for more than a few weeks at a time but this is what I normally see: 1.) how does your company handle auto scaling groups in AWS? When orion scans a cloud instance it will see the new server and automatically pick it up as a Cloud VM, but the question is…