cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

This product reached the end of support date on March 31, 2021.

High Availability implementation for AppMon

jalpeshs
Organizer

Hi Team,

Anybody implemented "High Availability" for AppMon?

I found "High Availability" for AppMon collector at the below link

https://community.dynatrace.com/community/display/...

But didn't found anything for AppMon server.

Please suggest.

Regards,

Jalpesh

17 REPLIES 17

alexzagor
Participant

It doesn't have helpful info regarding "High Availability" for dynatrace Server.

Hi Jalpesh,

Actually the link does have helpful info in that regard. In that post there is a reference to a description on how to accomplish HA for the Dynatrace Server. This link: https://community.dynatrace.com/community/display/...

Even though the layout and some pictures are missing, the most important parts are still there. In short: you set up 2 separate DT Server instances with a VIP/F5 in front that would detect an outage on one of the server. The VIP/F5 in front would then be responsible for starting the other instance. Config on both instances is being kept in sync using symbolic links to important config files which both instances can access from a NAS.

But that is just in short. I suggest you take a look at the attached links and example scripts.

KR,
Kristof

Hi Kristof,

We are getting Page Not Found error for that link .

Please suggest.

Regards,

Jalpesh Shelar

Just remove the dot at the end of the URL...

https://community.dynatrace.com/community/pages/viewpage.action?title=Dynatrace+HA+installation+in+practice&spaceKey=LEARN

Hi Roman,

Hard luck.It's still not working.

Below URL redirecting to Community Home Page

https://community.dynatrace.com/community/pages/vi...

Please suggest.

Reagrds,

Jalpesh Shelar

Seems some problem with access rights, a PDF of the page is here: dynatrace-ha-installation-in-practice-v2-20160518.pdf

Thanks Roman!

Unfortunatelly, PDF has been generated with no pictures.

😞

I know ... the pictures were generic avatar's so I didn't bother checking why they were not exported.

Hi, Roman,

Are you able to share the detail on are the data and config files that need to be accessible by both dynatrace server ?

[extract]

A shared mount was created using SAN storage, this will allows us to share the data between the
two servers. I used the mount point
but this can really be anything
/data/dynatrace
as long as it is shared between the two server

Thanks a lot. will try this using Windows Server Failover cluster instead. Thanks

@Roman Spitzbart and @Kristof Renders, may we ask you to export that article available on the knowledge base portal to PDF and attach it to our thread?

It would be very handful.

vladislav_samoy
Inactive

AFAIK dT does not have HA for server out of the box. Collector does, but not server part. Yet. Need to find others way to solve this.

wongcs
Inactive

Can we use Microsoft Windows Server Failover Clustering ? anyone done this before using WSFC to provide HA to dynatrace appmon server ?

michael_schoet1
Participant

I'm also looking currently in the HA implementation scripts. Now looking in our 6.3 environment there are some files in server/conf-path, which i assume may be important, but does not get copied by the script, e.g. infrastructure.config.xml. Who knows if the script are valid for 6.3 installations? Is there an updated script available?

Michael

Matt_Bernard
Dynatrace Helper
Dynatrace Helper

Hi Gents,

One of my key customer configured HA in respect of our documentations mentioned above and reported that collectors have to be restarted once the VIP switched traffic direction to Server 'B' instead of Server 'A'.

The message "Collector Not
Connected" is displayed within 'Deployment Health\Incident' screen from AppMon Client connected to 'Server B'.

My customer confirmed that Monitoring is OK since collectors have been restarted but is looking for a more relevant solution:

  1. Automate collectors restart once VIP switched traffic direction
  2. They are currently using a script executed every 10 min which is looking for a particular string within collector log files which highlights that a restart is needed.

I am not really sure if option #1 is doable but is there a better approach for option #2?

Any help would be very much appreciated.

Thanks,

Matt