What is the proper way to switch to DR for robots?
Currently, we have to manually get into every unattended machine to manually turn off PRD and then log into every DR machine to turn it on.
Is there perhaps a better way to do so? Having over a 100+ bots makes it time consuming, and can take hours when other departments spent less than 30mins for a switch.
Is there perhaps a system on BPM to automate failover for robots? Or some other way?
Thanks!
@KhaledA7 Are you using the RPA Service? If you are not using it along with Pega Robot Manager, then you really need to be. It is the mechanism by which you can control your RPA workforce without having to login to each one. As for Disaster Recovery, I am not exactly sure what you are changing on each bot, but your first step is certainly to begin using Pega Robot Manager and the RPA Service to control your workforce and then from there determine what you wish to happen when bots are failing.
Hi Thomas,
We are using the RPA Service and Robot Manager. However, in case of a planned DR activity, we have to manually get on every Prod bot, exit the application, disable the RPA Service. Then go on DR bots, exact copies of PRD, and manually enable the RPA Service and start it, then start the bots.
The Robot Manager comes into play when we manually go on it and put the bots in a “hanger” work group where they don’t receive anything.
Are we using it incorrectly perhaps?
@KhaledA7 You shouldn’t have to manually do anything if you are using Robot Manager. You should be able to disable the schedule for your bots and then stop them from Robot Manager (you can’t stop a bot if schedules are enabled). Since the schedules are disabled, the bots won’t do anything once they are stopped. They will remain stopped until you either manually start them from Robot Manager, or at the next interval once you re-enable the schedule.
In my understanding of what you are doing, you want to have these extra bots sitting idle for the majority of the time. In that case, you would need them assigned to a work group that had no work assigned and then move them to the work group with work once you switched over to them.
You should be able to do all of this from Robot Manager without manually going into any bots.
From what I understand from my colleagues, the reason we do it this way is because it consumes time to do it from the Robot Manager. The interface is very slow to load, and we are waiting for an update from the BPM team in the company.
The update should make it possible to do as you said as it would be less time consuming. Per personal experience, the current Robot Manager is very slow loading pages or showing search results. It can take minutes to return the results of a search query and we only have a hundred and so bots.
And your understanding is correct, and we already do that using the scheduler, where we move work groups out of working hours. We do this manually during DR activities as an extra precautionary step in case a bot was still running or started by mistake.
Thank you.
@KhaledA7 I am not aware of any performance issues with any versions of Robot Manager when properly resourced and configured. I would suggest ensuring you are running the latest version and ensuring your Pega Platform is sized accordingly. If you need assistance with that, please open a support request and one of our techs can examine your environment to determine why you are experiencing performance issues.
Oh yeah, it is not something inherent with pega. From what I heard, it is indeed a resource and configuration issue.
There is already an update in dev that is very responsive. However, if I understand correctly, the process of updating requires a lot of testing and bureaucratic approval, especially in a bank. So, it might take some time.