Multiple Run Time Resource

[KeerthanaJ]

Hello,

I designed three process and i would like to allocate each process to a run time resource. 

Like

Process 1 -> Run time Resource 1 -  loads the item to queue 1
Process 2 -> Run time Resource 2  - process the item and load to queue 2
Process 3 -> Run time Resource 3  - checks and verifies from queue 2

So here each process runs in half an hour difference. 

So i need some suggestion - 

If it is a good idea to use 3 run time resource, where each bot executes a process?
Or to use a single run time resource which executes all the process? 

What are the real benefits of using 3 run time resource, when you can just use a single bot and execute all the process? I really want to know what makes the difference? 

Thanks in advance..







------------------------------
Keerthana J
------------------------------

[shashank.kumar280]

Hi Keerthana,

For this scenario I would consider the number of cases that needs to be processed by Process 2 and if there is a SLA associated with it. If there is significant number of cases with tight SLA, that would make sense to run the process across multiple VMs. (so as to meet the SLA)

For Process 1 and 3 if you don't see a overlap in terms of the timing of the run with Process 2 then it can very well be run on the same VMs.

The advanced work queues guide lays out some of the design consideratiosn for work queues https://portal.blueprism.com/system/files/documents/Blue%20Prism%20-%20Advanced%20Work%20Queues%20Guide%20%286.2.1%29%28EN%29_1.pdf 




------------------------------
Shashank Kumar
DX Integrations Partner Consultant
Blue Prism
Singapore
+6581326707
------------------------------

[AnkitSingh]

Hi Keerthana,

While Shashank's answer to check SLA and number of cases is crucial you should also check the average time it takes to complete a process. If loading item to queue takes around a minute for instance and you don't have huge number of cases to be uploaded you can have bot on resource1 to upload cases for yesterday and afterwards work on verifying cases from queue 2!

------------------------------
Ankit Singh
------------------------------

[John__Carter · Accepted solution]

To add the answers already provided, splitting a solution into different processes might also simplify the design, facilitate team (ie quicker) development, enable an incremental delivery and make maintenance easier. I repeat 'might' - it's not a guarantee and each project has to be considered separately.

The main price to pay is increased license utilisation, so any 'one time' savings during delivery could be wiped out by inefficient license consumption during ongoing BAU. Scheduling separate processes is also usually more complex than for a single 'all in one' process.

If you are running a process every 30 minutes, then ask yourself what the resource is being used for between those runs? If the answer is 'nothing', then perhaps there is a more efficient schedule that would still meet business requirements, ie more infrequent but longer runs. And when discussing requirements with the client, instead of asking them 'when is the soonest you need the work done', instead ask 'when is the latest you need it done by'.

------------------------------
John Carter
Professional Services
Blue Prism
------------------------------

[KeerthanaJ]

I got it. Thanks for the reply Shashank.

------------------------------
Keerthana J
------------------------------

[KeerthanaJ]

Thanks for the reply Ankit.

------------------------------
Keerthana J
------------------------------

[KeerthanaJ]

Hey John,

Thanks for the reply. I got a better understanding now on the advantages and dis-advantages of using multiple runtime resources.

------------------------------
Keerthana J
------------------------------