savell

Nice work - looking forward to the custom property enhancements and the F5 support. Also great to see the UnDP polling and retention setting changes - we have a number of pollers that don't need regualar updating (serial numbers and IMEI data for instance that we use for inventory reasons). This will help reduce…

Perfect - thanks Mav!

Yes please! This is getting to be a real problem for service providers like us. When you have several (many) multiple poller implementations - it is really time consuming and error prone with so many components across many servers. I would love a certified package type of approach - single package install with all…

The provided report "Devices that do not respond to SNMP" uses the field "SNMP Level" rather than "SNMP Status". SNMP Level is 0 if the device has never responded, however the report will not be accurate if the device once responded and for some reason (i.e. firewall of access list change) no longer responds. Dave.

Hmmm, not on my implementation (which is why I asked for it to be added - I wonder why I don't see it) Anyone else not see this field in the list? Dave. 

Forgot to go back and update this post sorry. Just to confirm that sp2 sorted out this issue for us... Dave.

Just be a little careful making this change. We have tried this previously () resulting in a failure after a short period of time. Will be very interested in your results. Dave. 

We have the same serious HP Proliant pain - caused by HP's decision to only provide this infrastructure information (i.e. physical disk status, fan, temp etc) when using the HP CIM/SIM providers. It wouldn't be because they are trying to sell the HP Systems Insight Manager product would it? I'm hoping that SolarWinds will…

Hi Sam, I opened case #61428 yesterday - the developers are work on it. Dave.

Here is another way to report this data (using a query in conjunction with a stored proceedure to do the pivot). This way you don't need to know the PollerID's - it will show you all that have been assigned. One small problem - we don't really use custom pollers, so my testing of this query is limited (although I sure…

Hi Christine, I think the term “isn't ideal” falls short here.I would think that a majority of your user base would not monitor all ports. Even hardware and software licence issues aside, there is no way I would want to monitor our tens of thousands of floor ports. These are however exactly the ports I would like the user…

Thanks for that Tim! (and thanks to both Josh and yourself for the quick replies). Dave.

OK, SP3 has been running on the additional web server for the online day, and dllhost.exe has grown from 75Meg to 191Meg. Doesn't look like this has had any impact - will re-queue the support ticket. Dave.

Applied SP4 tonight, and can confirm that our additional web server issue (accessing APM from the additional web server) has now been resolved. Haven't closed the ticket yet - but will do do in the morning...(thanks to both the support and development teams!) Sav.

I couldn't load up our smaller SLX implemenation enough to reflect the I/O load numbers that our primary system is seeing. I guess I will have to wait until I'm ready to do the v10 upgrade on the primary system. On a positive note, the v10rc2 install on our smaller system went very well. No issues at this stage, and a…

Thought I would drop in my upgrade feedback... We have the luxury of a test/dev Orion implementation with only a few hundred nodes that we updated last week just after sp1 was available. This was a trouble free and easy upgrade. However, once we had MIB tables available and the web based node management options - it was…

OK - we had the same issue as documented above. Have applied the workaround as suggested by Sham - the service now starts. We did have NCM 5.0 working fine - this issue only showed up for us when the server was re-booted and the service failed to start. Thanks for the workaround Sham - I assume a fix will be in the next…

Just a note, in our case the custom poller memory leak and the dllhost.exe memory growth issue are seperate issues. My reasoning.... the dllhost issue appears on both our primary SLX machine (that also runs the custom poller), and also on our additional web server machine (which runs no other Orion processes). Dave.

 OK, support ticket requested I install SP3 - which I was finally able to do last night on all our systems (I just love our change control system). No issues with the SP3 install - should know in about 12 hours if this has made a difference. Dave. 

Well, it certainly resolved my slow inventory issue - the job now completes in 1 hour 21 minutes (the last completed run before the index change took 24 1/2 hours). This is for almost 1700 devices and 20 concurrent inventory processes. It is even faster than pre version 5.5 - although it's not fair to compare, as I have…

Thanks Josh - however you can't change this setting while the Hot-standby server is in standby (you can't connect to the polling engine). Any suggestions? Dave.

Thanks g@z - I see where these have been added now! Dave.

That's great news Michal. I have a good opportunity to do some benchmark tests between 9.5.1 and 10(rc2). Some be able to have some comparison numbers on Monday. Dave.

I see this perhaps once or twice a day, although the NOC screens see it multiple times (I'm assuming just because they are refreshing 24x7). Dave.

I think I 'may' have found an answer for my slow inventory issue. Refer Will be able to confirm once a full inventory run completes overnight. Dave.

Have been advised that this issue has now been identified, no fix eta just yet. Dave. 

Hi Michal, I use Process Monitor (http://technet.microsoft.com/en-us/sysinternals/bb896645.aspx) - Microsoft provided tool for windows monitoring. Dave.

Great - thanks Casey. While a pain, at least it's something we can live with for the time being. Dave.

Have tried reducing this down to 200ms and after running for an hour the inventory has only completed a handful of devices. I'll open up a ticket for this one. Dave.

Thanks ET - I'll open a ticket. I'm collecting topology data ok - it just seems that the collector is polling nodes for OID's that don't exist on a subset of our devices (and resulting in these errors). I thought it could be related to how long the topoplogy collection and calculation is taking. Dave.