Proposed SCSM 2012 day to day workflow for operators

  • Thursday, May 03, 2012 10:39 AM
     
     

    Hi,

    We're trying out SCSM 2012 as a helpdesk for the IT Department, I've got it all installed and so far it's looking very promising. However I've got some questions about how others use this as a helpdesk system...

    For the end user the SharePoint portal give an excellent simplistic view on how to raise a ticket and view your open requests etc. Which is great we need that for our users :-)

    But I'm a little stumped on the operators experience.

    It seems you really need the console installed as an analyst, which is fine, but looking at Work Items and running through some test requests, no flags or pop-ups appears as new requests come in. The console doesn't even seem to refresh itself if you're in a location like 'All Incidents'. So it seems you have to set up mail notification to let analysts know that a request has come in.

    But does that mean analysts are generally living in email, and only going into the console when they get a notification? Which begs the question if thqat's the case why isn't the exchange connector built-in. Or are analysts expected to regularly switch between the views they cover to make sure nothing has been missed?

    I got to point out that I'm very new to this, and I've probably missed something here, and I'd really like to hear how others are managing the requests as they come in. Also any quick methods they've found for analysts themselves to raise a request from a phone call or similar.

    Many Thanks,

    Andrew

All Replies

  • Thursday, May 03, 2012 12:47 PM
     
     

    Hi Andrew,
    The views don't auto refresh as when you have an installation that might have a few thousand incidents the performance impact of refreshing automatically could be severe.

    Usually people will scope the views down to better reflect their internal processes and structures as well as using the queues and groups.  Then it's a case of training analysts to refresh every so often, or like you say setup notifications to prompt them to go back to the console, but I wouldn't recommend this for dedicated Service Desk analysts as they would get swampped with e-mails potentially.

    You can use templates to help enter in common information when manually raising an incident from a phone call.
    There's these couple of options too:
    http://blogs.technet.com/b/servicemanager/archive/2011/09/08/incident-quick-templates.aspx
    http://contoso.se/blog/?p=1827

    Cheers,
    Steve


    My System Center Blog

  • Thursday, May 03, 2012 1:30 PM
     
     

    Hi Steve,

    Thanks for your reply, I understand it's a difficult question to answer as every company will have a different need. Here we have a team of 6 in the IT department, ranging from front-line technicians, to server side technicians. It's a tricky one as we all have our areas, for me I'm mainly working on our virtual environment, which includes system center. But when we have to we all get stuck in to the day to day tasks. We're certainly not ever going to see thousands of requests come in though.

    So it's case of how do I pitch Service Manager with the least amount of grumbles from the guys. Do I go down the email notification route, and probably have closure buttons and the like, with the ability to pop-in to the console, or the service desk analyst method of learning to refresh the console. I think for us the former is probably a better suit. I need to look more at custom views, queues, and groups.

    Would have been nice to have the option of auto refresh, or flags on views, for when new stuff comes in, with the ability to reduce the rate or turn it off if it becomes a problem.

    If anyone has a similar size team to us and is actively using SCSM, please let me know how you're getting on and what works for you..

    Andrew

  • Tuesday, August 14, 2012 5:57 PM
     
     Proposed

    HI Andrew - Looks like we are in the same boat. Smaller company and have been testing SCSM for the last 3 months making sure the config is similar to what we had. The e-mail notification is the only we we have it set. We have e-mails funneled through a specific e-mail group and have everyone in that group to get a copy (currently..will change once in place since they don't need two copies). Well the way I've set things is that they get a notification when a request has come in telling them that they have been assigned incident. The template has the title, description of the incident so that they really don't need the original e-mail. If there is a file attachment then yes they will need to review the attachment using service manager.

    I have also setup workflows based on our previous setup to apply templates based on the description words so that we can automate the procedure a bit more. The only thing is a few might slip by depending on how good you have your users trained to type in stuff in the body of the e-mail.  The the notification goes out to the group about the incident and they can go work on it. It seems to work just fine and even if the incident is assigned manually they get the notification just fine.

    I am also implementing the change management part of it for tracking of changes on our internal software....financial stuff...for auditing purposes. Not sure how far along you got but I can certainly describe more if you want me too.

    Thanks 


    Help

    • Proposed As Answer by needhelp..... _ Thursday, September 13, 2012 8:46 PM
    •  
  • Friday, September 14, 2012 7:28 AM
     
     

    Thanks for all the replies,

    It's been a few months now, and to be honest we've started moving away from emails. The team mainly use the console now, and users are getting OK with the SSP. Although due to the nature of SCSM, i.e. the very poor link between analyst and affected user and lack of console refresh, emails are inevitable. But if you're not careful they quickly lead to spamming!

    @NeedHelp... like you say we have also worked on the look of our emails, and now have a very well laid out email template which provides the core information to the team for new and resolved jobs. The rest is done within the console.

    We have also been running a Runbook which notifies different users on assignment and re-assignment based on criteria, as this is not possible out of the box. Worked brilliantly, and if you have any requirements for complex notification workflows (especially based on relationship changes) I highly recommend it.


    Andrew France - http://andrewsprivatecloud.wordpress.com

  • Friday, September 14, 2012 12:48 PM
     
     

    Andrew, thanks for your posts, they've been very useful as I try and get a grasp on Microsoft Service Manager. I am currently test driving Footprints for a new ticket system (currently on SDE) but just found out we are also inquiring Microsoft System Manager suite, which includes Service Manager. For you and anyone else reading, two questions:

    1. Andrew the Runbook you are talking about, what exactly is this - 3rd party software or advanced configuration of SM?

    2. Those who have a medium size company and have moved to Service Manager from another ticketing system, what do you think? If you could roll back clocks, would you do it again?

    3. Any feedback, especially from people who migrated from an existing ticket system would be awesome!

    Cheers and thanks in advance,
    Lauren~

  • Friday, September 14, 2012 1:04 PM
     
     

    Good morning Lauren -

    2. The system can be configured in any way you'd like but they make you work to do alot of it. From what I can tell alot of 3rd party software might already have the majority of options you need to customize or setup in Service Manager. The integration with other Microsoft products is great but the work needed to get the options you need/require is a bit crazy.

    Also due to the fact that they always seem to find things wrong we've delayed our rollout until Oct 1st since the console is a memory hog. I've patched a few computers(called microsoft) which seems to helped but I could not afford to roll things out if my analyst where going to spend more time waiting for the console then working. I am a fan of Microsoft but they seem to never get it right until after the first or second service pack....or what they call roll ups(outs) now.  Being this is their 2nd version of Service Manager I was hoping for a bit more.....but we will make the best of it.

    Thanks....(sorry for venting)


    Help

  • Friday, September 14, 2012 2:51 PM
     
     

    Hey Lauren,

    Runbooks are created in System Center Orchestrator, another product in the System Center suite, and essentially it takes SCSM from being quite a frustrating service desk system, to having the potential of processing things automatically in a way that I've not seen any helpdesk come close to.

    Take 3 scenarios as an example:

    AD user group membership, something that can change quite often (especially here!), and has a time overhead on the team. You can create a Service Request in SCSM, with all the questions needed to carry out the task, that uses a template which has a Runbook Automation Activity template. This Runbook Automation Activity would fire off a Runbook in Orchestrator, and it is essentially a series of powershell commands in a work flow (no coding necessary). This Runbook would then gather all the information in the SR and make the changes to the AD user groups membership, with no interaction needed from the team. You can of course add an approval activity :-)

    Analysts resolve tickets, Affected users close them, well that's the theory in practice they never do. So we have a Runbook which runs once a night that automatically closes any resolved tickets that are over 2 days old.

    System Center Operations Manager is great for checking the health of your systems, as long as you've tamed it! However the remediation side can be limited. Now we can have SCOM raise a ticket in SCSM when a particular fault occurs on a server, this ticket can have a Runbook assigned to it which specifically deals with the fault, and runs through any remediation processes we need. Once complete the ticket is closed. All logged and no-one needs to be woken up in the night!

    This really skims over what you can do with Orchestrator Runbooks with SCSM, and the whole System Center suite, but hopefully give you an idea.

    I can't honestly say we have come from another ticket system to this, but we have had 4 spectacular failures in the past, and since then just relied on email. This one has stuck, and the guys are using. It does have it's problems, and I've raised enough of them here and on Connect, but it works and once we start building into it it should start saving us time. It is Microsoft and it is System Center, so there is an element of hair loss and total confusion, but that's been par for the course in my experience of the product since SMS came out. But there always loads of people here willing to help :-)

    Hope that helps.


    Andrew France - http://andrewsprivatecloud.wordpress.com