fraser-sky

Thanks for the suggestion, we already get emails for the backup jobs and can see if jobs havent run by scanning the emails, i want to remove this manual check part for an automatic alert 

Ended up raising another support case and they now believe this to be a bug, no recommendations from support as yet

thanks for that, was hoping it would be possible but seems we wont be limiting admin connections

the script runs on the server it is assigned to so assume it will be local execution, so ignoring the fact there are 2 different return messages, 1 saying no quiet devices and the other swql error, on the 2 different tests, when running the monitor it does not display any of these we will be upgrading to latest release…

it doesnt have to be the local admin account, we do have a domain admin account, whichever way is possible or simpler

Thanks for reply You are correct, the exact scenario we are looking for

Thanks for the reply when i am testing from edit script, this is my output which is valid and correct and output of the Component monitor test - not sure why that is displaying something different but it is a valid output But the SAM summary show no values which is the same as the DB Solarwind support advise its the agent…

Thanks this put me on the right road, for some reason ${VirtualServerID} displayed all the virtual servers but ${ID} worked, only displaying the data for the specific VIP

That is a good idea having a separate server but not something we have at the moment, when you say SAM scripts can create alerts independent of the the status , How would i go about doing this ??

Thanks This was the initial alert we were using but once it failed over and if we were going to be running active on the other side for a period of time we had to manually adjust the alert criteria, which i am now trying to do automatically, does look like i m going to have to script it with SDK

Issue i have there is if we are working on them the nodes will be muted so no alert will get generated so wont have an alert action,

Thanks for the quick response, i was hoping for something within solarwinds to make the change, and using sdk to update the DB as last resort ( just in case i break something)

Found a table entry for Nexus devices that have cli polling enabled Orion.Nexus.VirtualPortChannel table has a LastSync entry, using this to get an early indication that cli polling is failing

I worked around this issue and go direct to the device to check the last commit date

Hi KMSigma Used the Cirrus entries and get the same result, running the query in SWQL returns the results I am after but running the query from python script on the Poller returns blank. I am trying to get the devices that config has changed on so i can run the rest of the script to export the config backup file to a…

We are running 2020.2.6 Testing the query in python and it seems it is the NCM. bit that is breaking it if i only use Orion.Nodes it returns results """SELECT TOP 10 Nodes.Displayname FROM Orion.Nodes as Nodes""" But using only any of the ncm tables it returns nothing """SELECT TOP 10 CoreNodeID FROM NCM.Nodes""" or…

Thanks for your input I used your format and i get the same result ======= RESTART: C:/Users/xxxx/Desktop/Python tests/Test-main-query.py ====== DisplayName >>> original python script ##omitted connection to server## So both queries run but i only see results from engineid, i dont see any results to extract from cfgchange…

Thanks for the reply But even using that format AND ModifiedTime >=ADDDATE('HOUR',-72,GETUTCDATE()) I get the same outcome, the query returns results when running it in swql studio but when running from python script results are blank

Managed to figure something out got SAM monitor to run swql query to get the last Modified time 1st and only run file backup if it has changed within last day my query if it helps anyone SELECT ModifiedTime, ConfigType FROM NCM.ConfigArchive as c JOIN NCM.Nodes as n on n.NodeID = c.NodeID WHERE SysName = 'xxxxx' AND…

Thanks to you both, I have the alert working Not sure if there is a more SQL efficient way to run query but it works this is the adjusted query SELECT Nodes.Uri, Nodes.DisplayName FROM Orion.Nodes AS Nodes left Join ( Select NodeID, CurrentValue from Orion.NPM.CustomPollerAssignment as P WHERE (CustomPollerOid =…

Hi yaquaholic Thanks for the reply I should have mentioned I am a network engineer and sql is not my thing , it took me forever to get alert working in swql. So not sure what you mean by pull values into alert.

Found the Table in swql database Orion.NPM.CustomPollerAssignment

Thank you all for your suggestions I did look at Exit codes but couldnt get my python script to work with them but managed to use the Statistic to set the level of the monitor which worked. But I have changed things a bit now so only a Statistic=3 will change status of Monitor to warning. So I have an alert that triggers…

Thanks again that is the correct syntax print(results['results'][0]['EngineID'])

Gave that a try and now get error: print(results[0]['EngineID']) KeyError: 0 I get the right output when using : for row in results['results']: print("{EngineID}".format(**row) But just want the single output as I want to use it as a variable in the script. swql query code = results = swis.query("SELECT EngineID FROM…

Hi Christycooper That worked, I am getting the query output now Thanks

Thanks for the reply I have found my mistake, I was connecting to the wrong swql server in my pythn script so my result was right as there was no entry ( insert face palm emoji) I do have another question though , the result will only ever produce a single line, how would i get it to print only the Engine ID nad not the…

Okay so it is hard coded into Solarwinds/NCM and cannot be changed.

Thanks Adam That is exactly what I was after. I can also see where i went wrong, Missed the i after Orion.NPm INterfaces, Did not Join Nodes and and missed the i after WHERE. I have learnt a little more about sql but think will leave it to the experts and post another question when I get stuck. Thanks Again

Found the reason it was not seeing the command prompt. SSH window length is default (80), the full command prompt is 80 characters so the F5 creates a new line to put the cursor on, NCM sees the new line and not the cmd prompt. Workaround is to adjust the prompt timeout and use virtual prompt. If anybody knows where in…