Service 'forgetting' Workflow and BPM?

[lomixture]

Hi there -

We are currently having an intermittent issue where the Catalogue items within Services 'forget' the flow they are supposed to be attached to. Similarly, we've also had it within progressive captures where the 'Switch Capture' option forgets where it's supposed to point to.

Is this something others have experienced, or is it instance specific?

Thanks

[Hornbill Staff DR]

Hi Lauren,

 thanks for your post.

It sounds like you are experiencing two different issues:

"....intermittent issue where the Catalogue items within Services 'forget' the flow they are supposed to be attached to."
I think I need a bit more clarification on this one.
1) Do you mean that tickets are being successfully raised against a service but sometimes the tickets don't have a BPM associated (i.e. there is no green progress bar visible at the top of a ticket)
OR
2) When you view the configuration of a catalog item, the two fields where you specify a Pro Cap flow or BPM are empty?
OR
3) Something different

"Similarly, we've also had it within progressive captures where the 'Switch Capture' option forgets where it's supposed to point to".
In relation to this one, do you experience a perpetually spinning Hornbill icon during progressive capture? This typically occurs when a progressive capture flow reaches the Switch Capture node but it cannot identify the flow that it must switch to in order to continue the progressive capture experience.
Here, it is necessary to understand how you are determining which flow to switch to. Is it based on the selection of a catalog item, or is it specifically referencing the name of another flow? If its the former (which is usually how the out-of-box flows are set up) can you confirm that a catalog item is being selected by the analyst each time?

Thanks,

Dan

 

[lomixture]

Hi @DanielRi, thanks for the quick reply!

Quote

"...intermittent issue where the Catalogue items within Services 'forget' the flow they are supposed to be attached to."

I think I need a bit more clarification on this one.
1) Do you mean that tickets are being successfully raised against a service but sometimes the tickets don't have a BPM associated (i.e. there is no green progress bar visible at the top of a ticket)
OR
2) When you view the configuration of a catalog item, the two fields where you specify a Pro Cap flow or BPM are empty?
OR
3) Something different

This is currently #2, when looking at the catalog item, the two fields are empty. I had to refill every single catalog item with the correct Pro Cap flow and BPM.

Quote

"Similarly, we've also had it within progressive captures where the 'Switch Capture' option forgets where it's supposed to point to".

In relation to this one, do you experience a perpetually spinning Hornbill icon during progressive capture? This typically occurs when a progressive capture flow reaches the Switch Capture node but it cannot identify the flow that it must switch to in order to continue the progressive capture experience.
Here, it is necessary to understand how you are determining which flow to switch to. Is it based on the selection of a catalog item, or is it specifically referencing the name of another flow? If its the former (which is usually how the out-of-box flows are set up) can you confirm that a catalog item is being selected by the analyst each time?

This is in relation to the Switch Capture node losing the information of which Pro Cap to switch to, the box is empty within the node. This then leads to the spinning Hornbill icon because it doesn't know where it's heading.

In our instance, we have built a 'Selection' Pro Cap flow which is used solely to choose the category and redirect to a new Pro Cap flow for that specific category.

This sporadically loses the secondary flow information and makes it a pain because all of a sudden it doesn't work!

Lauren

[Hornbill Staff DR]

Hi Lauren,

 thanks for the update.

If I may consider the blank fields in the catalog item configuration, I've experienced something very similar when working on a customer site recently. However, in my case, after a few seconds the pro cap and bpm that i'd originally configured would appear in the fields. This was, at times, a good 2 - 4 seconds delay but it seems to be an issue loading the information rather than anything wrong with what has been set and stored. 

Are you able to confirm that when this occurs the pro cap and BPM values do eventually appear? How frequently are you experiencing this?


On the topic of the Switch Capture node, values will certainly disappear if the flow specified is deleted. Aside from that there shouldn't be any reason why this would happen. As the node stores the ID of the flow behind the scenes, renaming the flow is not an issue.
Are you able to give me an example of the flow that was specified in the node and then disappeared? I would suggest monitoring the situation and gathering more information on the frequency at which it occurs. Is there a particular point in time when this started happening?

Thanks,

Dan

[lomixture]

@stepghg for info

[lomixture]

41 minutes ago, DanielRi said:

Hi Lauren,

 thanks for the update.

If I may consider the blank fields in the catalog item configuration, I've experienced something very similar when working on a customer site recently. However, in my case, after a few seconds the pro cap and bpm that i'd originally configured would appear in the fields. This was, at times, a good 2 - 4 seconds delay but it seems to be an issue loading the information rather than anything wrong with what has been set and stored. 

Are you able to confirm that when this occurs the pro cap and BPM values do eventually appear? How frequently are you experiencing this?


On the topic of the Switch Capture node, values will certainly disappear if the flow specified is deleted. Aside from that there shouldn't be any reason why this would happen. As the node stores the ID of the flow behind the scenes, renaming the flow is not an issue.
Are you able to give me an example of the flow that was specified in the node and then disappeared? I would suggest monitoring the situation and gathering more information on the frequency at which it occurs. Is there a particular point in time when this started happening?

Thanks,

Dan

Dan,

Unfortunately not, this is not a case of it not appearing on the screen, the only reason we recognised this issue is because a few test jobs we logged went against the generic workflow rather than the specific ones. We also kept getting this error message in the bottom corner, and this was also experienced by @Lauren yesterday when testing together. I've emailed the SwitchOn email account about this today.

SR00000367 is an example of this which @Laurenlogged when testing yesterday.

The flows have not been renamed or deleted in the Pro Cap flow.

Within the KEPSM instance, it is the Facilities - Fault Reporting - Selection, and Facilities - Service Request - Selection where we are having the issue. This has happened as recently as this morning on the 'Electrical' switch capture in the Fault Reporting. This has previously been raised with @Bob Dickinson when he was working with us, but as happened more recently and frequently since.

Thanks

 

[Hornbill Staff DR]

Hi Lauren,

 My previous advice would not apply to this message you have now reported ("The flow is poorly formed - loop detected...") and it sheds new light on your situation. 

There are underlying checks in the progressive capture flow switch-capture mechanism that protects against a flow referencing itself and creating a loop. It may not be literally referencing itself during the switch, but I suspect that if the configurations of two flows are very similar then maybe the system is interpreting it as a possible loop. Perhaps this mechanism is being over cautious in its evaluation,  in which case we may have to seek developer assistance in order to "tune" this for a future product build.

Anyway, before we get to that stage it will be necessary for me to understand the way you have your progressive capture flows set up. As your organisation has purchased a Success Plan, and we are going to require more detailed specifics in relation to your configuration, I'd recommend that we move this investigation into an incident logged with support. I can see that you have already reached out to us via email however we are still waiting on confirmation of your supported contacts that we hold on record before we can raise an incident in your name.

For the time being, I will raise the incident in the name of one of the other supported contacts we have on record and someone will be in touch later this afternoon.

Thanks,
Dan

[lomixture]

Hi Dan,

The email is being sent in imminently with updated contact details for Support contacts.

Thanks,
Lauren