RichardLetts ✭✭✭✭✭

shutdown the solarwinds server processes during the upgrade. You'll loose stats collection for that time period but you won't end up with: a) spurious alerts (very bad) b) incorrect down-time data for your network c) quicker than unmanaging and re-managing everything.

Syslog messages have the advantage that they are generally unformatted text, but that is also their downfall in that they are then hard to analyse. Using a product like SPLUNK to aggregate logfiles is a good idea, you can save yourself hours of grepping through files by using something like that. Splunk has a free…

No, but if you get is working let me know. I asked for SAML support... let me see... Coming up on FIVE years ago: Jun 28, 2012: aside from that you would have to get integrated IIS authentication working, and then have that pass the right stuff into Solarwinds, and hope that works. I've not had time to play with this to…

I'm interested in this, and also for the Traps table. You would also have to update the database maintenance tasks to take advantage of this. /RjL

there is not a limit, however, if ONE node in the set doesn't respond as expected it will blow the whole discovery process. I have a custom property that I set after importing nodes into UDT. I then: a) pick nodes with that property cleared b) change the group by to machine type c) clear the scanning on individual nodes

The interfaces [implicitly] depend on the node, which was down, and so the interface alerts were all suppressed. yep, you need to add an alert for the node.

from SolarWinds Network Performance Monitor Release Notes Interface Discovery Improved with Filtering * Exclude virtual interfaces and access ports, or specific interface type * Use pattern-matching, including regular expressions, to select interfaces for monitoring * Filtering applies to interface name, description, VLAN…

Personally, I'd be using the SNMP trap received to search the trap for the string you want to alert on and setting the color there; going this in the alert manager requires substringing (which means it has to non-index scan all of the data in the table rather than doing an index-scan looking for an integer value) I've no…

wait till the end to do the NTA upgrades. If you look at the releases notes for each product they tell you the minimum version you can upgrade from -- I will skip versions if I can e.g. NTA 2.2.1 ⇒ NTA 3.0 ⇒ NTA 3.1 ⇒ NTA 3.7 ⇒ NTA 3.9 ⇒ NTA 3.11 ⇒ NTA 4.1 so you don't need to install NTA 4.0 Orion Network Configuration…

the device/node monitoring will probably sort itself out. the interface monitoring will probably be all messed up -> there is no way to tell solarwinds the SNMP ifindex and help with remapping interfaces in the UI. -> will have to remove the unknown interfaces & add the new ones

Are there any error messages in the solarwinds trap receiver logfile? see: Solarwinds NPM 11 SNMP Trap Error in Logs Are there rules for your trap received causing it to drop inbound packets?

Transmit discards you say... are you sure you are not trying to put 1Gbps down a 100Mbps port? This normally happens when the sending switch has traffic backed up on an interface that is already full of outbound traffic. In this case I'd not suspect a VLAN mismatch since it's the sending switch that is discarding the…

Is the SolarWinds alerting service allowed to interact with the desktop? if not, I don't think notepad.exe will appear

there is an out of the box alert for this:

Tip: following on from what mesverrum​ mentioned the backup jobs can general quite a few transient failures on a large network, and you might want to add one or more cleanup jobs that retry the backups for nodes that failed and email the result of that to someone to look at. to do this you create a backup job with a…

use sudo(8) to allow the user to run hdparm as root (but nothing else.) Allow A Normal User To Run Commands As root Under Linux / UNIX Operating Systems if you don't have sudo installed, then you should... it is great for allowing one to reduce the number of people allowed access to the root shell. Richard

monitor the switch interface that's plugged into it?

Is there any progress on this -- it is getting to the point where I cannot use the UnDP to add custom pollers to nodes -- with several thousand nodes of any particular type modifying a UnDP times out.

there is a standard MIB for PoE support -- I think I put in an enhancement request for this to be added to UDT in the mean time you can use a UnDP to putt the values back into the database, where you can then do something with them (draw graphs, generate alerts)

I use a custom SQL Alert on the nodes: Where [status] in ('1' , '3' ) and unmanaged=0 and DATEDIFF(MINUTE,LastSystemUpTimePollUtc,GETDATE()) >60 basically -- if the sysuptime is not being polled in the last 60 minutes throw this alert. This catches the SNMP agent dying, the SNMP community being changed on me, or someone…

I feel this... when you're running it for the 10th time during an upgrade.

4. Establish a tiger team for the Solarwinds server migration project for a day a) do the best you can you have the new nodes reachable by your new server before the day i. shutdown your old app server for maintenance ii. use the Solarwinds Upgrade buddy on your new server to monitor the up/down state of your servers, see…

here is what I have discovered as the minimum to get a node added in perl. The SDK may show minimums, but the webUI needs more than those to prevent giving errors in random places [aside: these should really have been put into the database as defaults] my $AddedNode=$swis->Create('Orion.Nodes', {'Caption'=>$ip,…

if you view the error report locally you will get some useful text that might indicate what has gone wrong

Probably not. e.g. 10.5.5 Use file-integrity monitoring or change-detection software on logs to ensure that existing log data cannot be changed without generating alerts (although new data being added should not cause an alert). I could write a SQL update statement that removed/rewrote entries from the logs being stored in…

NET-SNMP commonly found on linux systems does (you have to walk all three contexts separately).

I'd write a small perl or powershell script using the API to update the value Here is an (untested) perl version, takes an input file of nodeid and ip addresses (note the use of ip2guid to set that as well otherwise things will not work as you expect ) #!/usr/bin/perluse SOAP::Transport::HTTP;use XML::Simple;use…

What are you trying to do? You need to be a little more specific on what you are trying to achieve not the path you are following. In SQL studio this works: SELECT AlertDefID, ActiveObject, ObjectType, State, WorkingState, ObjectName, AlertMessage, TriggerTimeStamp, TriggerTimeOffset, TriggerCount, ResetTimeStamp,…

We're moving away from iperf testing to Y.1564 : http://www.exfo.com/solutions/metro-core-networks/bu2-carrier-ethernet/ethersam for commissioning carrier Ethernet circuits. This generates a report that the carriers don't normally argue the meaning of. Oftentimes they've discovered some mid-hop segment that has different…

create two alerts -- 1. if the Primary_Admin is populated, and use that to create the Incident 2. if Primary_Admin in not populated then create an incident record with the Operations team as the assignment group