Ramkumara11
Enthusiast
Enthusiast

Synthetic monitoring for Airwatch

Team,

On premise Airwatch guys!

WE are on premise (Version 19.07) and our servers often fail with lots of performance RAM/CPU issues. WE have 50k devices enrolled (4 DS and 2 console server)

We are having to reboot our VM's daily and do frustrating IIS resets.

Queries below:

Is there a tool which sends out sms or email notifications if there are performance issue seen?

Also, when the private Que goes over a certain limit, can we automatically trigger a reboot?

Kindly advise,              

Labels (1)
12 Replies
LukeDC
Expert
Expert

We use SolarWinds. You can probably setup some fancy monitoring using it. Or if you are so inclined use some powershell scripts to check these things as well.

Also seen Nagios - The Industry Standard In IT Infrastructure Monitoring  before and that was nice as well.

HimanshuMishra
Enthusiast
Enthusiast

Hi,

I am experiencing similar issues. We are also On-prem(1907) and have over 31K devices with 2 DS and 2 CS. IIS worker process kept using more than 95% CPU and memory. High usage choked the hub app. It would not launch. IIS resets and server reboots helped but did not fix the issue. Vmware support could not find the root cause but suggested to add more CPU and memory. That fixed the IIS issue for us. No spikes in last 3 weeks. We are working on adding another DS and probably update to 2001 soon.

-Himanshu

0 Kudos
Ramkumara11
Enthusiast
Enthusiast

Thanks LuKeDC.

What does the tool solar-winds and powershell script basically monitor?

Do you check things like IIS worker processes, CPU/ram utilization etc?

Ram         

0 Kudos
LukeDC
Expert
Expert

You can monitor any of these things using a proper tool. Just have to configure it.

LukeDC
Expert
Expert

Look at the architecture guides and make sure you are spec'ing your servers properly. It really makes a difference. Also make sure Windows is configured to handle the load with Virtual Memory settings, etc. If these are virtual servers (VMware, azure, etc) make sure they are also tuned to work with high pressure situations.

Introduction to Workspace ONE UEM Recommended Architecture

LukeDC
Expert
Expert

0 Kudos
Ramkumara11
Enthusiast
Enthusiast

Thanks Luke, they are spec ed well. no issues on that.

infact we are over sized

what configurations have u done on the solar winds side? that is something i d love to look at.

0 Kudos
LukeDC
Expert
Expert

Not a whole lot. Just basics. CPU usage, disk usage, etc. Haven't found a need to go much further as I haven't seen performance issues. Are you using Cache servers? Memcached is recommended for large deployments.

Ramkumara11
Enthusiast
Enthusiast

No i am actually exploring MEM cached.

Will it help in reducing cpu/ram spikes?              

0 Kudos
sturmanc
Enthusiast
Enthusiast

We also had memory leak issues with previous versions of Workspace ONE (AirWatch).

To protect the service we implemented IIS App Pool memory limits.  This will recycle the IIS App Pool and prevent server memory saturation. Maybe also look a CPU limit also.

This is a MS SharePoint link but gives you the instructions.

Application pools recycle when memory limits are exceeded (SharePoint Server) | Microsoft Docs

Also check your logging levels.   If you have any always set to Verbose that will cause CPU/Memory issues as the log files are generated and written.

We have 35k devices with 4 DS and 2 CS servers.  Each server is Windows 2016 VM with 6 CPU/12 GB RAM.  We don't see any performance issues on 19.09.07.

Ramkumara11
Enthusiast
Enthusiast

Hey Sturmanc,

Thanks for the info. What are the typical value u have set for V-ram?

WE have 4 DS and 2 console and 3 ACC servers.              

0 Kudos
sturmanc
Enthusiast
Enthusiast

The DS App Pool we set the limit at 5GB because that is the one with the memory leak before.  But it should be a percentage of your max RAM.

0 Kudos