MrSampsonite

Do you know how to just put the caption drive letter in reports? I can't see a way to use the SQL filter query in the formatting. Thanks!

I know this is old, but any progress? Looking at the MIB from Orion via UnDP it doesn't look like there are much else to look for. 

Lots of examples. Anything really in SQL. We store versions of the software, SQL has it's own version info, patches, service packs, etc. There is info about the servers (OS, CPU types, etc). That can be found with other tools, but one central place would be great. Also we store information in text files in a certain path…

Couldn't figure out a clean way to export and import. So I just grabbed the IPs via SQL query and then pasted them into a new discovery based on list of IPs. And in there I could create a new SNMP credential string. Thanks!

Thanks. I'll give it a try.

Hi Sorry, I think there's a misunderstanding. I'm talking about getting those green/yellow/red bars into the report that I display in an Orion View. Though getting them into a pdf or email report would be great, that's not really what I'm asking for. I'm also not trying to take the canned report SW provides (the bottom of…

HI I was talking more for any graph. For example we are monitoring several perfmon counters on our servers. For some there are spikes so high that it makes reading the normal values impossible. If we could adjust the Y axis to be just above where the average utilization is we could probably make out what they are visually…

Thanks, except that doesn't have an XML entry for multiple UnDP or Interfaces that I can see. 

the "fixed" comment was custom sorry. The subject line of all alerts we have start with "ALERT:" and then the real subject line. When it's reset or no longer an issue we start it with "FIXED". Just a little visual queue when we're looking at the subject line. It really means 'RESET' but we were going for normal lingo, not…

Yes, it's APM. The trigger says 1. when node status is not DOWN 2. when component name is 'a' or 'b' or 'c', etc (which I copied the name from) 3. Component status is DOWN 4. Application name not 'other application i don't want to trigger alerts from' The property type is APM Componnent. Reset when not true. No alert…

thanks. is there any way to customize the timeout? even for all types?

We do have APM with many perfmon counters polling. So are people recommending this setup: Server1: IIS and Orion APP Server2: Orion Poller Server3: Orion DB

thanks. i'm opening a support ticket.

did you ever find a fix for this? we're at the latest version now in 2011 and it still doesn't work for our entourage users.

I think I can defer for different components if I make multiple alerts and the filter only applies to certain components. Would that work?

custom html is a way to get a static view, but it is not a long term solution. there's no way to change what you're looking at quickly. you have to into each piece of custom html and change the details individually. this takes forever. plus it's not nodeid filterable, so it makes it unreliable across nodes.

Thanks. I mean admins shouldn't be doing a TON of stuff so I doubt the log would get out of control. I'm still a bit surprised they're using log files too. Just store this stuff in the database. Much easier to manager and query, sort, etc that way.

Are you referring to a script specifically that you know of? So maybe we put the content out to a text file or even into a SQL database. Then have APM query that file/table periodically and store in Orion?

forgive me if this is wrong, but is it essentially saying 64 bit counters on faster media (like gigabit NICs) are faster but will return the same info needed?

Thanks. Is that different than a feature request? Is this more of a "bug" issue?

Thanks. I guess I'll have to see what the cost differences are. Right now we only have up/down alerts for them so I might be told by higher ups to just do " / " and nothing else.

CPU tends to vary much more than memory. I'm not concerned about a few megabytes here and there, but if one says 1MB consistently and the other method says 300MB consistently, we're not talking about fluxuations. We're talking about measuring different things. For a memory starved system, knowing who is stealing it all is…

I don't recall setting or changing that length. So I assume it's the default. Is that somewhere in the NPM settings screens? I assume that when a node is added, somewhere in sql there's a time stamp even if the event record is not kept.

what does it mean to be "less than 'doesn't exist'"? not sure I follow you.

Must be a bug. I'm using IE 8, Firefox 3.6.2 and neither of them show any box like you show. I tried creating another one, checking the other radio buttons first, and still nothing. This is my SW versions: SolarWinds Orion Core 2011.1.0 SP1, APM 4.0.2 SP3, NPM 10.1.2, IVIM 1.1.1

Thanks. Unfortunately that doesn't really help because we have last XX events on many different views depending on what we're monitoring. we don't really spend much time in the message center. maybe we should?

We don't use traps really and the only thing that monitors our servers via SNMP is Orion. I'll have to check the path next time I see it.

I've never created one before. Can you point me to an article on how to do this? And do the traps check for free space often and will Orion alert as soon as a trap comes in? Not sure how that all works.

looks like I edit the graph details on each node in system manager, but I don't see how to do it for all of them. there are so many nodes and so many settings it would take forever to go to each graph on each node and customize each setting.

thanks. that seems to work. however it must be that this nic has never been up since it returns no results for it but does for other devices. the part that's confusing is that this is a linux machine, and it lists the loopback adapter as UP and the eth0 as down. but the server is pingable.