If there was a bunch of changes to the CMDB what would you recommend we do to go back to an out of the box states but keep or edit the data in a way to conform to the out of the box solution.
Even though this class doesn't cover CMDB deeply, we have been getting a lot of questions about CMDB "health." So, I thought I'd put this link here for reference:
"CMDB Health provides a framework for configuring CI remediation. Remediation lets you proactively apply corrective actions to unhealthy CIs in a managed and standardized fashion."
A question was raised about multiple locations, here's how we handle it - For users that work from home or visit multiple sites or travel a lot: we flag those mobile users and ask their location on the REQ/INC portal.
Different customers use different approaches to normalization. Some use the "older" method of normalization (i.e. normalizing specific fields chosen by them), others use data services (as you indicated above). Both work well for different companies. One golden rule, however does apply. DON'T USE BOTH. As you can imagine, one approach will try to override the other and vise versa.
I notice you use manufacturer name as a prefix on asset tags. Is this best practice and can you explain why? We currently just use a numeric system ...next available number.
We do that primarily in the labs as a way to reinforce what assets we are dealing with. There is no "best practice" naming convention we are showing here.
Are asset_classes (tables) part of the CMDB? It looks like it is a "separate" database as all asset classes start with ALM? Second question (sorry) -> What does ALM stand for?
We just talked about Model Categories where either CI Class or Asset Class is empty - my question is why would such a record exist in the first place, since the Model Category record is used to map CI and Asset - so if one is empty, then there is no mapping so why would such a record exist in the first place?
Will we cover Model Management where we use the same Model globally but they are country specific? The issue is that we cannot move some Models between some neighboring countries for legal reasons, e.g. disposal rules, import/export regulations, but other Models can be shipped across borders to use existing inventory.
Procurement application allows to "receive" the assets, and as a result - to automatically create the new assets in ServiceNow. In addition, once PO is ordered, assets can be created in SN of before they are received.
How would I model a PC which has a VM instance of a server? As the VMware is independent of the PC model, would it make sense to create a model to indicate both?
PC is a hardware asset, therefore, it would have an asset record. VM may not be considered as an asset but rather a CI and would not have an asset record. However, it would have a CI and be associated with the machine that is running on.
We discovered 5.3 million CI's and the bandwidth was negligible - for us. Things might be different for your infrastructure, We found that getting buy-in from your network admins was key. We had our network teams run network tools running over small subnets of non-critical CIs to see the impact before going live.
Assets may be transferred from one stockroom to another. Once the transfer order is created and the assets are added to the TO, the substate automatically changes to Pending transfer.
Pre-allocated assets are assets that do not have asset tags yet, no allocations or financial responsibility. Those assets are/can be stored in multiple quantities. This is the only substate for hardware asset that allows the quantity to be changed once the model is entered. The Pre-allocated record can be split or individual allocations made.
Would like to know this as well, as we have our distributor that labels our assets for us and we currently use the serial as the tag number. So the serial and tag field are the same, any issues with this?
Asset tag is not a mandatory field and does not have to be used. If at some point the organization decides to asset tag all assets, it may require some data clean up/replacement vs. populating the field with the newly available asset tags.
I don't see this as an issue at all. My example given is just a consideration. It is not uncommon for organizations to have various formats of Asset Tags.
Is there an ootb way to create a CI record from an existing asset record? For example during our go live event we did not import retired CI's from our old system but the asset record was still listed as in use. Now we have assets with no CI. If we create the CI it will attempt to create new asset record.
If you go through the "asset route"--creating an asset record and then creating a CI record--where do you specify which CI record to generate in cases where an asset class links up to more than one CI class?
Can we manage BYOD CI creation where the own device matches a company model - e.g. Apple Iphone is a company device that is bought for certain levels of management, but others can buy their own?
I don't have any hard data for this but I can tell you from my experience that it is very little. In China, network bandwidth is very expensive and new network demands are always tested. We ran Discovery for a range of 900 PCs and our China team reporting no noticeable impact to the network
Question on Discovery, if we manually create a Computer Asset / CI with a SN and then Discovery finds a computer with that SN, i am assuming that the Discovery will not create a duplicate CI but update the Computer Asset / CI with the relevant data from Discovery?
Asset Management and Configuration Management are two different processes with different purposes. By keeping those records separate we can run and maximize those processes.
