Hello fellow forum members,
i am using IPmonitor since 2000 for my company and our work is based on this tool by 60%. We have seen IPmonitor improving well during the years and upgrading to v9 caused some discussions because according to the whitepaper the improvments seemed not to legitimate the price increase. In addition we already suffered from the takeover of Deepmetrix (former Mediahouse the innovator of Ipmonitor) by Microsoft and the following stop of development on Livestats. So we lost our investment in this product. I was afraid this could happen with IPmonitor too. Finaly it was group decission to update because of the all new dashboard. This feature could replace our second application for live monitoring of snmp data (cpu, bandwidth, memory usage, temperature and so on).
The update process worked well and the IPmon v9 setup programm guided us through all the steps. After 30min (it takes some time for v9 to update the 8.5 monitors) we where done and curious how things would look now. And here the challenge starts.
1. All our monitor names had been messed up by replacing our custom monitor/device names with the fqdn or ip address of the device. E.g. we named a device "Ramses" (e.g. windows or linux server) and v9 renamed it to "ramses.ourdomain.net". Or a switch was called foundry-core-1 and renamed to 192.168.100.1 (the device has no dns name).
2. All monitor groups have been placed under a group called "updated-dateofinstall". The groups still contained our old monitors but expanding the group in a browser window takes up to 15 sec. and ends with a script error.
3. In "my network" we see many devices but not all monitors for one device have been placed under the device allthough they show up together in the former 8.5 group. Some monitors are now splitt over 3 devices allthough the former monitor name and/or ip address where identical. Now we see under one device many "foreign" monitors. This happend to all ADO monitors (MSSQL, MySQL)
After this first shock we started to sort things out. Creating new groups, renaming devices and monitors to make things look like we want (and need) it. And the challenge goes on:
1. While you can move devices to an other group you cant do that with monitors. This means all the ADO monitors sitting under the wrong device have to be created new under the device it belongs to. But this fails for the "ADO - User Experience (SQL Query)" wizard throwing 80004005 errors every time. For MSSQL this is due the fact v9 assumes "a trusted connection" but you cant select a different security mechanism in the wizard. For MySQL this is because you cant select "run this as external process" in the wizard. The only way is to use the "ADO SQL Connect" wizard but this is not what we want because this monitor does not query a database and compare the results.
2. Due to the fact searching through 380 devices to find out what its former name was (those now carring only an IP) we decided to create some devices new. We started the "add new device" wizard and entered the IP to scan. Surprisingly we found you cant disable snmp discovery in the wizard so it takes up to 3 min. until you get a result. The scan result page again shows the dns name or ip address as device name so we changed that to the custom name we want. Finishing the last wizard step ends in the device overview page but still the device and monitors carry the dns name or ip address. So we had to change all display names manualy (or with mass edit and substring replace).
3. While scanning devices we discovered that not always all ports got selected allthough the settings in the "smart wizard" have been set to activate e.g. ping, http, ftp, smtp if found. This adds a second step to modify the scan results.
4. Adding new devices and regrouping things gave us the opportunity to take advantage of the new "smart groups". In 8.5 we had e.g. a group called SLA for customers with service level agrement. This caused more work because after adding monitors to a group named like the device we had to add this group or monitors to the SLA group. This should be better now with v9 because you can create a smart group based on a tag value. So we added custom tags to devices like "SLA" with content "yes". Now we created a smart group called "SLA" and selected "add monitors where tag" hoping to find our "SLA" tag in the drop down list. But we didnt. Only scanid and createdate is available. We searched everywhere how to add tags to get them into the drop down box. No go. So we had to do it like in 8.5 without smart groups.
5. Finaly we added bandwidth monitors to the routers to take advantage of the new dashboard to get a "Live View". And again big surprise. You cant add a graph to the dashboard like in reports. This is only available for ready made gadgets like CPU. Well the main reason we updated to v9 is not given anymore.
6. Well so we had to create reports with a 5 min timeframe to get a kind of "Live View". This was like in 8.5 and caused no problems. We had to modify the time selection because we want to see the last 60min for a device. We selected "start this hour" and go backward 60mins. As a result the report is empty. Timezone is set to our timezone and the device clock is correct. After a while of playing with the selection we found to set "start this hour" go backward 3 hours, go forward 120mins and as a result we get a report showing the last and next hour. That cant be correct.
Conclusion: The update to v9 took 5 days until now and still we are not finished to fix the messed up installation. The new features are partly worthless and some old features have been broken. The new webinterface is much slower than in v8.5 and gives some headaches with different browsers (Firefox, IE). Solarwinds got it managed to make a good product worse but to take more money for it. We already opend a ticket and descriped all the problems (thanks Stephane for supporting us well as in the past 7 years) but for now none of the issues has been solved. Our decission to update was wrong and we will not renew our subscription until the problems are solved and the improvments are worth the higher price tag. This is the way how you can frighten off faithful customers.
All our monitor names had been messed up by replacing our custom monitor/device names with the fqdn or ip address of the device. All monitor groups have been placed under a group called "updated-dateofinstall".
All your monitor and existing group names have remained unchanged. Your previous relationships (Monitor-Group) has been preserved, and all the maintenance, alerting configuration, and v8.5 reports are intact.
The upgrade did manufacture new containers (devices) and name them based on your dns configuration to relate every Monitor to a single device. In this regard, the use of devices by the new features probably trumpet an invalid configuration in your eyes. See my comment just below about support helping to change that configuration.
Some monitors are now split over 3 devices although the former monitor name and/or ip address where identical. Now we see under one device many "foreign" monitors. This happened to all ADO monitors (MSSQL, MySQL).
Each Monitor can exist to only one device. Try not to think of a device like you would a Group. A Group can contain any number of Groups and Monitors. A Device is intended to represent the “one place” where a Monitor exists on your network. If the upgrade code couldn't figure out an actual address, it placed it within the "ipMonitor Host device".
With your open ticket, the support technician can assist you with moving the Monitors from one Device to another, tweaking all or just a few.
The groups still contained our old monitors but expanding the group in a browser window takes up to 15 sec. and ends with a script error.
We have a performance enhancement going through our QA department at the moment. If you've brought up the subject within a ticket, you'll be offered the performance patch (as it is now).
While scanning devices we discovered that not always all ports got selected although the settings in the "smart wizard" have been set to activate e.g. ping, http, ftp, smtp if found. This adds a second step to modify the scan results.
If the Monitor currently exists in your configuration, the Monitor won’t be recommended to be added.
We searched everywhere how to add tags to get them into the drop down box. No go. So we had to do it like in 8.5 without smart groups.
You can use tags for grouping Monitors, however you’ll have to put the tags on the actual Monitors. The way you describe it, it’ll be easier (at first) to making a regular Group and adding Devices in manually (if your intent is to group devices based on certain criteria). In 8.5 you could not Group ‘containers’ together.
You can’t add a graph to the dashboard like in reports. This is only available for ready made gadgets like CPU.
You can drill down into the Monitor itself to see your graphs update in real time on the dashboard. Admittedly, it isn’t like the reports interface you’re used to that’s really complex. The dashboard was intended to be (at first) simple / easy to use and configure.
I’m glad you have a solid relationship with Stephane. He’s awesome. I can see already that there’s some immediate movement to be made in your open ticket. This is a big release that unlocks lots of new organization potential.. I’m sure it took a while to organize ipMonitor the way you want it, changing all of that to maximize the benefit from the new features will take some too. Check in with Stephane (if he’s got your ticket) today.
thanks for your quick reply. As i already told Stephane this is the support we know from Mediahouse/Deepmetrix/Ipmonitor - fast, friendly and always helpfull.
I understand all what you wrote however needing a support technican to move monitors from a former 8.5 group into a device is a bit ugly and it would have been helpfull to simply have the same feature for monitors as you have for devices: tick the box and select the move to group icon.
About the dns resolving for devices i must repeat myself. We dont want it. We like to name things like we want and how it has been established in our company for 11 years now. It worked in 8.5 without it so why shouldnt it work in v9? A simply checkbox in the wizard "use dns" would solve it and people can decide what they want to use. Same for snmp scan. Some devices dont have common snmp tree and simply dont respond to a normal scan which causes time delays. Take 100 devices and each delays only for 60 sec in average and you have wasted more the 1.5 hours of expencive worktime.
About the tag issue and what you have said i am totaly confused now. Nearly any item can carry tags but you say only if a monitor has tags i can group them to smart groups? This would end up in a smart group containing monitors only and not devices. This would be the concept of 8.5 and does not reflect the device concept in v9. Dont understand me wrong i like the device thingy (except the dns resolution display name messup) and therefore the concept should be used consequently and not only in parts. Adding tags to all monitors would be more time consuming and would have useless redundant data. It is usefull if you can tag each monitor and it should prevail the same tag if set on a device (e.g. device sla = yes, monitor http sla = no, monitor cpu without tag; smart group sla contains devices with sla = yes; if cpu reaches limit alarm for smart group sla is trigered because device has sla; if webserver goes down no alert for smart group sla is triggered because monitor tag overrites device tag; this way you can easly group and alert your devices)
The drill down into the dashboard view was a good tip and if we could select a custom timeframe this would be a 1st step to what we want.
I will contact Stephane again and ask for the patches.
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process.