Transcript from CA Service Operations Insight Office Hours - Live Online Chat – May 5, 2016
Melissa Potvin (CA) to Everyone: Hi Everyone, we will get started in just a few minutes
bcarb to Everyone:Quick question about upgrading to CU1 from 4.0. Is there a best upgrade process that will result in the minimum amount of down time? We have Manager, UI server, IFW connector servers, and catalyst connector servers - not as concerned about reporting.
Jeff Morris to Everyone: @bcarb - There's not really a staged way to upgrade because of the dependencies, although there are a couple of things you can do
Jeff Morris to Everyone: @bcarb - You can shut down the UI server and upgrade it first, just don't restart it - i.e. makes sure "start services" is unchecked
Jeff Morris to Everyone: Then shut down connectors and manager
Jeff Morris to Everyone: Upgrade manager first and get it started re-initializing
Jeff Morris to Everyone:While this is happening you can upgrade the connectors
Jeff Morris to Everyone: When the MGR is back on-line (triage tests "completed") you can restart UI server and connectors
bcarb to Everyone: @jeff - thanks . that should minimze some of the disruption
Jeff Morris to Everyone: @bcarb - You're welcome. It's not a huge savings. That being said - we are looking at some ways to improve this and have some ideas so hopefully that will be on the roadmap soon
sergio to Everyone:Hi all. Are there SCOM, UIM and Spectrum connectores available to SOI4.0 CU1?
Joseph Poutre to Everyone:Any chance of an AutoSys connector in the future?
Jeff Morris to Everyone: @Sergio - Yes - connectors are generally independent of SOI versions
Jeff Morris to Everyone: @Joseph - We don't have a productized one, but there is a field-developed solution I could discuss with you if you're interested - feel free to shoot me an e-mail jeff.morris@ca.com
Joseph Poutre to Everyone: @Jeff - thank you.
bcarb to Everyone:We recently had to restart the SOI manager and all the connectors reecovered - except the UIM connector (where we had to restart the catalyst container containing the uim connector) , Our UIM connector is GA but we have not yet upgraded to CU1. Is that "reconnect" feature improved in CU1?
Jeff Morris to Everyone: @Bcarb - Brian, not sure there - all connectors should reconnect. From your description I can't tell if this is a connector issue or a container issue. If you had a support ticket on this, feel free to e-mail details to me and I can follow up with you
sergio to Everyone: We also have CAWA DE and we need to integrate it with SOI. Have you any way or any plan to do that?
bcarb to Everyone: We create tickets in SOI via service desk - but we also create tickets in UIM (also to service desk) - yet the alerts that UIM sends to SOI dont contain ticketing info as a standard out-of-the-box configuration. If other domain managers create tickets, wouldn't it make sense to see this in SOI (and have it integrated into the displays)? Is there any work in progress in this area?
Jeff Morris to Everyone: @sergio - We don't have that on our roadmap. We would have to follow-up with Product Management on this. I'm not as familiar with CAWA DE vs. AutoSys but if they are using the same back-end DB the solution I mentioned earlier might work - research would be needed for this
sergio to Everyone: Thanks @Jeff
Jeff Morris to Everyone @bcarb - Within our portfolio, if you're using SOI it is generally the preferred method for integrating into Service Desk, as this is a key use case around Centralized Alert Mananagement. That being said there may be something on the UIM connector roadmap for this - will have to follow up with Product Management on that
bcarb to Everyone: Thanks @jeff
bcarb to Everyone: An optimization question: if we have mulitiple event policies for a connector, does it matter how many we create, or would it be better to try to stuff everything into fewer policies? Also, do we know in what order the policies are processed?
Jeff Morris to Everyone: @bcarb - Regarding the sequencing - if the alert data matches criteria for multiple policies then the policies execute in alphabetical order based on policy name. Generally you would try to make your criteria mutually exclusive if possible, but if you need to run multiple policies for the same alert, you have to set the reevaluate flag and then be careful you don't create circular references in your policies
Jeff Morris to Everyone @bcarb - Regarding the optimization piece, will have to get back to you on that.
bcarb to Everyone: so if there is any chance that you want the event to be considered by more policies then you would set the reevalute flag. but the normalizations dont have this reevaluate. they are always considered?
Jeff Morris to Everyone @bcarb - It sounds like you have a specific use case in mind. Feel free to send me an e-mail and we can set up a quick call/webex to discuss
bcarb to Everyone:@jeff - k thx
Melissa Potvin (CA) to Everyone: Last Call for Final Questions..
Melissa Potvin (CA) to Everyone: Thanks for joining! Transcripts will post to the community (as usual). next event May 31 CA UIM Integration: Operational Mode. That concludes our session for the day. Have a great day everyone!