- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎01-04-2018 01:32 PM
Hello All
My company just got done implementing SN Discovery/CMDB with the help of an SN Partner Vendor. As the vendor roles off I need to identify who will be taking over the role of the Discover and CMDB Admin. I have 4 SN Admins who been admin to the tool when we first implemented about 2 years ago. They are sys admin can they do the role of a Discovery or CMDB Admin. What skill set is requested for these role and we implemented 4 CI how much time will be needed to spend on Discovery and CMDB
Any suggestions/help would greatly be appreciated
Solved! Go to Solution.
- Labels:
-
Discovery
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎01-05-2018 07:17 AM
I can give you a couple of ideas of day to day activities, but much of what you'll do will depend upon your situation. To quote Steve Jobs, "You can't connect the dots going forward. You can only connect them looking backwards".
Some daily activities:
1. Analyze the ECC Queue for errors with probes. Things to look for are
- SSH errors, may indicate that the discovery user cannot login to a device, or that the network is blocking access.
- Other errors, like permission errors. Does the discovery user have permission to run the privileged commands on each device? These errors will take on different forms depending upon Linux, Windows or some other device.
2. Review announcements from other groups which include lists of devices, such as servers, computers, printers, etc. Verify that the servers in the announcement are present in the CMDB.
3. Receive incidents from your users about the CMDB. Often, this is how you'll learn that a device is NOT being discovered.
4. Setup new discovery jobs as needed for items form #3.
5. Monitor the discovery jobs each day. Are they all completing? If not, find out why. Do they need more time? Should they run at a different time each day?
6. Periodically, check the load profile of the ECC queue. Are you overloading any particular time? Move discovery jobs around to smooth down the peaks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎01-04-2018 02:06 PM
Hello Kiran,
This is a great question and I think that there are numerous answers and which are best will:
A) depend upon your specific situation, and
B) change over time.
Here's some considerations:
1. How large is your CMDB? Specifically, how many CI's will you be interested in? Our CMDB has millions of CI's. But we are only interested in about 100,000 ( computers, installed software, disk/file systems, IP Addresses, printers, etc ).
2. How dynamic is your infrastructure? Does it change frequently? How often are significant CI's added, retired and replaced?
3. How customized is your CMDB? Will it need customizations in the future?
4. Have you restricted access to those tables and/or fields which are not important?
These considerations will give you an indication of what level of work will be needed. In our case, the CMDB is "usually" a part time job, taking up less than 10% of our time. But occasionally, we encounter an issue and the CMDB becomes a 110% job. (Yes, that means overtime). We also receive requests to extend the CMDB and this can result in a development effort to change Discovery to accommodate our new needs.
I recommend that all your admins have some familiarity with the CMDB with respect to the use cases that are important to your organization. But have at least one person develop the skill set to develop changes to both the CMDB and Discovery Software.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎01-04-2018 04:26 PM
Thank you so much on a daily bases what would a Discovery admin do, CMDB Admin and Discovery Admin role can it be combined or would you suggest it should/should not and sorry I am so confused someone coming in with only sys admin knowledge is this going to be a chalenge
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎01-05-2018 07:17 AM
I can give you a couple of ideas of day to day activities, but much of what you'll do will depend upon your situation. To quote Steve Jobs, "You can't connect the dots going forward. You can only connect them looking backwards".
Some daily activities:
1. Analyze the ECC Queue for errors with probes. Things to look for are
- SSH errors, may indicate that the discovery user cannot login to a device, or that the network is blocking access.
- Other errors, like permission errors. Does the discovery user have permission to run the privileged commands on each device? These errors will take on different forms depending upon Linux, Windows or some other device.
2. Review announcements from other groups which include lists of devices, such as servers, computers, printers, etc. Verify that the servers in the announcement are present in the CMDB.
3. Receive incidents from your users about the CMDB. Often, this is how you'll learn that a device is NOT being discovered.
4. Setup new discovery jobs as needed for items form #3.
5. Monitor the discovery jobs each day. Are they all completing? If not, find out why. Do they need more time? Should they run at a different time each day?
6. Periodically, check the load profile of the ECC queue. Are you overloading any particular time? Move discovery jobs around to smooth down the peaks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
‎12-20-2020 05:27 AM
Check out this video, it will clear all your doubts and help you to understand Discovery queries in details.
Link: https://www.youtube.com/watch?v=30JbWVsusyE&t=10s&ab_channel=ServiceNowHelpdesk
It help you to understand below points.
- Discovery Overview
- Discovery prerequisite
- Understanding Discovery Phases in details
- Discovery credentials and IP Affinity
- Mid Server Management with Cluster and Load Balancer
- Schedule jobs
- Set up discovery from scratch to end
- Live implementation with real world data.
- Troubleshooting on various aspects
- Many more other issue related to mid server, CIs
- Cloud discovery
- Service Mapping
Please mark reply as Helpful/Correct, if applicable. Thanks!!