Apologies if I seem to be starting a lot of discussions lately but we are getting pushed to get our last few defects fixed asap.
We have a 'Release Management' Process which wasn't working as the precondition test was incorrect - I've fixed this.
However, looking at the process as designed by LANDesk, I'm not sure if this design should actually work correctly.
Can you please take a look at the 2 preconditions that are followed by manual actions (diagram attached).
Are the 2 manual (windowless) actions 'Return to Release Being Scoped' a valid method to 'Move on' to another status within a process?
The process IMHO looks OK but maybe I'm missing something; difficult to see from just the picture.
Are the "Return To Release Being Scoped" in fact the same Action or 2 different actions but one with trailing space in the name to make it look visually the same in the process Action panel. I presume that if both preconditions are satisfied then only one instance of the Action is displayed in the Actions panel?
I wonder why they chose though to use 2 seperate preconditions rather than use a calculated condition that satisfies both tests in one go; "Any CNN or Release Component Removed".
You can't have 2 stacked preconditions for an action item. You can have 2 different preconditions for an action item. In the 1st case 2 things need to be true before the action item is available (this will not work). In the second case, either of the conditions can be true to allow the action item to be available.
Thanks for taking the time to replay Julian.
I need to check but I think that the requirement was for these to automatically move on without any manual actions taken – could it be that our consultant has simply forgotten to make these automated actions?
This makes a mockery of the UAT sessions we had before Xmas !!
Thanks for the reply too Carl.
The attached image was a bit ropey but there aren't 2 stacked preconditions - they are both in parallel under a single status.
The more I look at this the more it appears that the Process was badly designed and poorly tested - but we are 2 months into a LIVE system !!
Ah but in that case process designer wont let you have 2 preconditions feeding automatic actions from the same status because if both preconditions were true then the system would reach some deadlock state of not knowing which branch(es) to take. Will not allow you to save the process but I think it lets you design it still because it doesnt do it's dependency checking until save.
What you can do though is feed a single precondition (calculated of course) which evaluates to true if either of your originl (seperate) criteria is true and then feed than (via an automatic move on action) into a condition (which is setup to test for just one of the criteria) and then the YES/NO branches can go where you need them too. Think, without testing again now, I've done this in the past during process design.
Well, you live and learn Julian !
I've been using LANDesk for a year but haven't had any formal training so I'm learning as I go.
It makes you wonder what the consultants get paid for sometimes.
It now transpires that the customer wants to review the whole process which yet again points to poor testing in the first place.
Many thanks for your help and I'm sure I'll use your advice in the redesign.
Forgive me for striking a balance for my consultant collegues.
There are a couple of points here which you have to some extent referred to yourself. The first is that the testing that was done was not good enough and that isn't something that the consultant can manage. Second, it seems to me that the customer isn't entirely sure of what they actually want, and finally, on the whole, the consultants implement what the customer has asked for, yes there are implementation tricks and other techniques that they should apply as necessary, but it is the customers design at the end of the day.
I might also suggest, very nicely , that doing some training is not a bad thing. It is a complex tool and if you want to deploy complex solutions with it, then that deep level of understanding is essential.
Point well made Andy.
Apologies, this wasn't a cheap shot at any of our consultants as they were most helpful to me in the implementation phase.
I also agree that the lack of system testing and product training has made it really frustrating from a Support perspective.
I'll keep my views to myself in future.
Just wondering what training you would suggest for the likes of Keiron or myself? He seems to be in the same situation as me, self taught, has a good enough grasp on the product but doesn't have the deep level of understanding that you've mentioned.
I did recently do the Advanced Servicedesk training but apart from a couple of things everything else I'd already learnt from LANDesk documents or here on community. Speaking to others on the course who have completed most of the Servicedesk training courses they didn't think I'd benefit from any of those either. I'm currently "attempting" to teach myself BOO to be able to acheive more using calculations in processes but if you can recommend any in-house training I'd definitely be interested.
There is a 4 day administration course (LANDesk Service Desk Core Administration & Design) that I took was was very helpful. I would have been completely lost without it, this software isn't always as intuitive as we would like. They have a process course as well. If you know the admin basics and want to change processes (who doesn't) might be more what you want. I highly recommend the admin class for all admins. It is well worth the money as you won't need a consultant as much.
I found the consultant we had to be very capable and put together a nice starting point for us that worked well. No it wasn't prfect but it was pretty good. He did what we asked him to do, not what we wanted. Big difference there. He got us most of the way there and I "finished" the details. I say "finished" because after 2 years were still not done making changes but we do have a stable system that does what we need it to do. We just keep finding better ways to do things or new problems where Service Desk is the solution.
Thanks Carl I'll have a look at the Core Admin and Design course, see what it entails. I had hoped to do the Process Design course but after speaking to 2 people who have done it it didn't sound like I'd get much out of it. I had already built our Request processes from scratch pre upgrade to 7.4, our consultant did polish them up for me though, and even better showed me the bits I wasn't sure of. Since then he'd probably not recognise them if he came back over as they've been reworked and new processes created as well. As you say, you always find a better way of doing something - or in our case finding a way to stop certain analysts doing things!!
But definitely our consultant was great, I'd say in some things he didn't do what we asked for, he asked us what we were trying to acheive by it and then explained how we wouldn't get that with the design as we had it and reworked it so we would . And as I said, I was fairly new to being an admin of Servicedesk when he came over but he took the time to explain things to me and made me do them, and do them properly - I am now as pernickety about window design as he is