chisle · Employed! · ✭✭✭✭✭

I tried, but it didn't let me as it 'remembered' my user had already submitted it.

go to Mission Rules and select the week you want, then answer all the qeustion. Looks like they are now posting all days one week at a time. I just answered week 1 and 2 in one full swing.

Looks like they changed format. Go back to the mission rules, click on a 'week' and then follow the step-by-step prompts to fill in the questions.

Can we have an updated/comment from SW on whether these bumps are having an impact?

Got it. you seem to be using SAM much heavier than we do. WE find many lackings in SAM that we do workarounds for, and your suggestion is trying to mitigate that kind of work. So, I vote it up, and once in place, we'll use it, very likely.

AMEN! This would make us way more willing to deep-dive into the SAM structures....the fact that you have to call each one individually, and build multiple per host simply because you can only return a certain small set of variables to display i

on day 2, the exact name is 2 words, not just the one that it indicates. I copied them directly out of the text and pasted them. Though the one word the quiz was looking for was accurately in the answer I posted, it dinged me because I used the full phrase.

This has likewise casued some unwanted results for us. From time to time we end up with priority one calls regarding disk space for disks that were not part of the host when originally added. A report with actionable links would be great (add these, don't add those), or something to that effect. The ability to permanently…

I had a case where we submitted a feature request similar to this one. Not sure why the engineer didn't know this was already a request...it came up as a result of what we were tackling in the case that it would be good to know if a node were muted or not. Seems like a fairly straightforward change, don't you think?

In a simple deployment and test I found the agent numbers are way off as well. as an example: /PS --> 53830283264bytes, or 50.13 GB Agent reports: 47.58 GB That is a difference of 2.55 GB! /PS is 19% consumed Agent reports 18% consumed How, with a smaller reported overall size of the volume does the reported consumption…

Maybe not store all processes, but maybe the top N by cpu and/or by memory and/or by I/O would greatly reduce what is kept in the database, and also make it much more useful when drilling into a performance issue. We've used a monitoring tool in the past that allowed this kind of feature and were able to put the kabosh on…

If I understand you correctly, you are hoping that something would be created by where multiple templates could be applied to a server, and you could select which would override same variables? I.e., if template A has an active process limit of 100, but tempalte B has an active process limit of 150, you would like to tell…

Agreed. You can end up with a boatload of servers starving for resources but show as green because they respond to a ping. Not really a true health statement.

Yes, getting a 'feel' for what the end to end traffic should be so that we can deal with the one-off calls of "I think it's slow" and actually have facts to either validate or assuage the user.

6.8, and yes, I opened a case. I think we'll get it straightened out.

Yes, failures are where I'm going, as I need to see the attempt=failure code. this exists for nodes but not for components of nodes (cpu, disk, volumes, memory, etc).

Additionally, if certain values in the return data could be checked to make sure that the response is valid. I.e., you may get a page response from a server (other than 400), and it would appear to be "up," but since the database back-end is having issues, the data on the page is bad. So, if you know a good response will…

go to day 2...it is the same report...

Not good for audit when they find out that credentials for accounts that can do somewhat of a deep dive into a server/service are stored in SolarWinds...I would much rather that there were vault security.

We don't have a whole lot of HP-UX around here, but what we do have is poorly instrumented due to their snmp module not being very hardware savvy. Having an agent would really help get this up to snuff.

Did anyone on this thread implement linuxatemyram successfully? I implemented it, and have no data to show on my node screens. The data is poll-able (tested before deployed), but now, the cpu and memory graph is completely empty, and data ceased to be collected since the date/time I switched to using linuxatemyram. I left…

There is a widget at the node level. However, it is at the component within a node that we need this alerting, as there are many alerts that are created at the node level.

all those calls that result in 'run this sql delete statement' or 'add this to the database,' etc. , etc. could be preempted with something that is digging along. How about monitoring whether e-mails are going out or not? or depth of queue? Or load overall?

either the monitor could be configured with 'accept this many response pairs' or the script itself could dynamically supply that value. For instance, a file system monitor for *nix that actually lists the user-observed values of mounte file systems (as to actual space free), could have as few as one response pair, but on…

Absolutely a great idea. If it is already available, it beats me where you would 'find' it. We find in our alerts we can use hyperlinks if the property already exists within NPM/SAM; but if it isn't there, it is, as you say, copy/paste to get there. thank you.

If someone is able to figure out how to augment the net-snmp poller to pull this data (inodes) and report on it, that would be swell.

And add HP-UX? We've got some heavy hitters on that OS as well.

Yes!

LOVE This idea! Anything to get alerts more readily configured is a good change.

HP's out-of-the-box snmp module only has networking and basics. No file system, for instance. It contains the following: IF-MIB RFC1213-MIB IP-MIB TCP-MIB UDP-MIB IPV6-UDP-MIB EtherLike-MIB SNMPv2-MIB (see below) IPV6-MIB IPv6-ICMP-MIB There is no hardware/OS stuff in those MIBs