marcrobinson · Observability Detective · ✭✭✭✭✭

When the solarwinds server reports down, rdp to the polling server and attempt to load the web page from that server. This can be common pain point - the users can access a resource, but the polling servers were blocked. Ifyou have NPM, netpath to the url:port can be used to validate what solarwinds is experiencing. If the…

That event 4001 is one that I threw in a SAM template. If it trips a lot, then I know that sql connectivity is bunk. Also, Netpath in NPM is underrated. I have setup a netpath agent on each poller and main poller to port 1433 of the sql server. That will help eliminate traffic latency etc. There was an article somewhere…

@"KMSigma.SWI" Ok, that is a great way to find and relate which log files you need. I never thought of abusing, um using, the log adjuster tool for that purpose. I also agree that there is no singular document that lays out all the log files. There are some documents that do point out which log files are relevant to an…

I think I found the relevant KB article. I can confirm that I cannot edit the DHCP or SNMP settings. Importing from spreadsheet is another matter. The API may be able to update them too - But I have not validated that. https://documentation.solarwinds.com/en/success_center/ipam/content/ipam-edit-ip-address-properties.htm…

Can you share a bit more detail? Does ICMP (uptime) polling work? You mentioned three separate polling methods. Are there more nodes affected? Here are some links to get you started: Article Orion Platform Troubleshooting - Polling (site.com) Troubleshoot nodes that are unknown

I dont recall kiwi having HA. BUT if you have a load balancer that would be perfect. I had a kiwi farm behind a load balancer, and could reboot a kiwi server without causing havoc. Alternatively, logstash is also a good forwarder. 

I have noticed that you have a couple of threads open for database space, polling issues, and this one for maintenance. I suggest creating a support ticket as you may have some bigger issues to work with. Additionally, you can run the Active Diagnostics and it will give you some heads up. IT is located in the SolarWinds…

Try this to get you started. The orion.npm.custompoller**** entities should have the custom poller statistics that you want. SELECT TOP 100 cpon.AssignmentName, cpon.NodeID, n.Caption, cpon.RawStatus, cpon.Status FROM Orion.NPM.CustomPollerStatusOnNode cpon JOIN orion.nodes n on n.NodeID = cpon.NodeID

Just from the comment about a large number of alerts, I trust that the web console alert export/import process will be slow if not painful. Check these out: (+) Import Multiple Alert Rules into Alert Manager - Forum - Network Performance Monitor (NPM) - THWACK (solarwinds.com) Vinay.By and MesVerrum have suggestions/links…

@"vinay.by" and @"mesverrum" have the whole of it. I have run into situations where the IIS log had the details to determine if there were specific user actions involved. Monitoring IIS with SAM is also helpful to let you know the general performance of a given webserver. Otherwise the Message center and the Audit…

 VIM and VMAN both use the VMware api to talk to Vcenter. I would check the VMAN logs for connection errors. You should see data much sooner than a couple of days. At least 12 hours as @"Seashore" indicated. 

I agree with @"stuartd" , and you can enable session trace logs in NCM which have more detail. Generate session trace logs for NCM and NPM modules (site.com) I think It may catch errors before the node was entirely setup in NCM. But definitely useful to troubleshoot connection errors when NCM is enabled on a node. 

PowerShell, APIs, PowerBI and get more comfortable with VS Code. 

Link to the migration guide docs: SolarWinds Platform Migration Guide

I am not sure how to feel about an in place upgrade. I understand that it works much better now from teammates. It is a matter of probably running config wizards to 'refresh' the solarwinds installation on those servers. Otherwise, there is an option to migrate to new servers using the HA functionality to hop over by…

SolarWinds is working on the issue with the support documents.

Recreating my post. Also noticed that other pages are not loading. Adding screenshots. One of those pages was working about ten minutes ago. 

An orion map based on groups might be an alternative also. Depending on your platform version, I believe 2023.4.x is where that ability was released to add orion maps to modern dashboards.

Your second part of the response can be done via the following - list of modern dashboards with links. Generating the status - I would consider creating dynamic groups around your studio custom property, then showing the group status for the first column. Cleaner table. There are other ways, such as only listing down nodes…

@"mesverrum" Thank you, I was not thinking about that, just looking at how to get this work in SWQL. @"3wal" Try adding this to the second query's where statement: and ss.Role >1Looks like Role = 1 matches the masternumber in the npm.switchstack entity. 

This is an ugly answer. I was able to use subqueries, an arithmetic function, top function and converting the result to numeric. The top and numeric conversion I cannot explain other than I received errors without multiplying by 1.0 and 4000 repeats without the Top1. I changed your queries to values that I could test, but…

Remove everything after the AND statement and add the following: ORDER BY r.RecordTime ASC That should return all the records available and sort by time ascending. That will eliminate issues with the last 30 days.

Looks like my last reply did not send. I added a date comparison to the where clause to get only the last 30 days. Try this: SELECT op.OperationName, r.OperationInstanceID, r.RecordTime, r.MinRoundTripTime, r.AvgRoundTripTime, r.MaxRoundTripTime FROM Orion.IpSla.OperationResultsHourly AS r JOIN Orion.IpSla.Operations AS op…

Try something like this, which uses the hourly entity and the operations entity: SELECT TOP 1000 op.OperationName , r.OperationInstanceID , r.RecordTime , r.MinRoundTripTime , r.AvgRoundTripTime , r.MaxRoundTripTime FROM Orion.IpSla.OperationResultsHourly AS r INNER JOIN Orion.IpSla.Operations AS op ON…

@"adam.beedell" Adam is too kind in his assessment of the performance during variable loading. Hateful comes to mind when waiting on variables to load when you search. But the fact that VIM returns very little in this case is even worse. 

@"cheryl1" Hey Cheryl, this is something to ponder as you really need to work backwards on some of the vim entities. They work but you really need to watch how you link them. Rachel's initial approach works if you can resolve the hostid, but you cannot as it is written. I was surprised that I had to work differently to get…

 @"adam.beedell" I was looking at the following and noticed that multiples could be returned. Top 1 eliminates that. SELECT TOP 1 H.DataCenter.VCenter.displayname FROM Orion.VIM.HOSTS H WHERE H.DataStores.DataStoreID = ${N=SwisEntity;M=DataStoreID} The other entity that looks good based on the earlier thread are those…

Taking a quick peek and hoping that the slight shift in entity for the alert (changing from virtual host to virtual datastore) is not the cause. This can cause some interesting and subtle changes in formatting the swql variable to retrieve the value. SAM application and component alerts are a great example, as you slightly…

same here. going to poke solarwinds on this one. Thank you for pointing this out. 

Just updating this thread. The ticket is still moving forward. It was a bug that was resolved in the past. The issue occurs when ASA's report data from dropped connections apparently. They can report negative data, and then this data becomes a bigger issue when it is scooped up. Sounds like the resolution may be a new fix…