With Andrew Nam
Welcome to the Cisco Support Community Ask the Expert conversation. This is an opportunity to learn and ask questions about Cisco Unified Computing System (UCS) Director with Cisco expert Andrew Nam.
Cisco UCS Director was designed to operationally integrate bare-metal and virtual data center infrastructure resources to address complex, time-consuming, manual, and compartmentalized management processes. These processes burden IT organizations, preventing them from achieving business agility and efficiency. Cisco expert Andrew Nam will provide an update on installation, configuration, and troubleshooting VM provisioning process using Cisco UCS Director.
Andrew Nam is a data center solution engineer in the DC Solution team in Sydney, Australia, responsible for orchestrating the end-to-end solution support of Cisco Data Centre solutions, including Cisco UCS, Cisco Nexus architecture, VBlock/FlexPod, VDI/VXI, and cloud solutions. His areas of expertise include routing and switching, load balancer, WAN optimization, VPN, and firewalls. Andrew has worked for Cisco for more than 13 years and has 15 years of experience in the networking industry. He graduated from New South Wales University in Australia with a mechanical/manufacturing engineering degree and holds R&S CCIE 9586, VMware VPC5, and Citrix CCA - Xendesktop5 certifications.
Remember to use the rating system to let Andrew know if you have received an adequate response.
Andrew might not be able to answer each question due to the volume expected during this event. Remember that you can continue the conversation in Data Center community, sub-community, Unified Computing discussion forum shortly after the event. This event lasts through January 17, 2014. Visit this forum often to view responses to your questions and the questions of other community members.
How do I collect log files when I start troubleshooting the UCS director? thanks for your help.
From UCS Director main menu go to "Administration" > "System Administration" and then look for "Support Information" tab.
Click "System Information and Logs Link Page" and it will bring you the "System Information and Logs Link Page" as below.
Click "Download All Logs" button and it will save the log name as Latestlog.zip to your local pc.
I hope this answers your question.
Thank you for the answer, Andrew. I really appreciate it. Another question I have regarding this topic is if there are performance issues how do I go about checking it? Looking forward to your answer.
Thank you again,
No problem. Thank you for the particiation.
Going back to your follow up question, UCS director is the SW tool for Bare metal provisioning and Virtual provisioning so the performance issue is mostly visible from the lack of response in the Web site when you need to monitor or perform the provisoning task.
Symptoms often exhibit as the initial login page would display fine however the loading of subsequent pages and navigation through the site is not appropriate. It would take too long to load a new page once it was clicked on.
Here is the checklist :
1. VM resource reservation
Try to reserve the resources for CUIC Virtual Machine via vSphere :
CPU --> 2000 Mhz
Memory --> 3GB
Note : This requirement may vary as per the new release is available.
2. Infrastructure Polling interval
Increase "Infrastructure Polling interval" from default value to 60 min.
This setting can be configured at Administration > Physical Accounts > Infrastructure System parameters
3. Monitoring property
Only enable property absolutely needed. We have seen some customer has enabled all the properties in "Advance Configuration", it can significantly contribute the slowness on the system.
You can adjust this setting at Administration > System Administration > Advanced Configuration
Note : As a course of troubleshooting, untick "Performance monitoring" or untick everything and monitor if the problem goes away. Enable property absolutely needed.
4. Map Report
Review the map report to see if there is any VM in orverload status. If there is any VM in overload status, you need to increase the resource to the VM.
The Map Report is found from Virtual > "Map Reports" tab
5. If none of them help, open TAC service request and forward all logs and system info (Adv).
I'm so glad to see we have this session about UCSD. We have deployed Cisco Virtual Network Services in our DC with multi-tenancy. We aim for no more than 100 tenants as a start and we hope UCSD is able to help us to achieve orchestration in deploying individual tenant environment.
Attached are the workflow for an initial setup, and a workflow for onboarding a new client. Would you be able to show me how the workflow can be related to task library and what are the essential steps to do this in UCSD?
We have installed CUCSD_BMA_4_1_0_0_GA.
Many thanks in advance,
This Ask the Expert session may not be the right place to show you all the essential steps for how the workflow can be related to task library to achieve your initial setup flow chart, and onboarding a new client.
However, I can briefly walktthrough "Client blade Provisioning" task so it might give some idea and how you can go about.
Assumption : the infrastructure used in the following examples consists of:
- Vmware VCenter
- Cisco UCS
- NetApp ONTAP controlled storage
To be able to provision Cisco UCS blade server in automated fashion, you need to create and define the below UCS entities beforehand.
- Create UCS Organisation
- Create UUID Pool
- Create MAC Address Pool
- Create WWNN Pool
- Create WWPN Pool
- Create vHBA Templates
- Create vNIC Templates
- Create UCS Policy vHBAs
- Create UCS Policy vNICs
- Create Storage Policy
- Create Network Policy
- Create SAN Boot Policy
- Create LAN Boot Policy
Once you create all the policy above, you are good to set up a workflow container for the client blade provisoining.
1. Create UCS Service Profile
- Add a ‘Create UCS Service Profile’ workflow task and select ‘Map to User Input’ for ‘Service Profile Name’. Select the ‘Service Profile Name’ dropdown as created when the workflow container was created.
- ‘Create UCS Service Profile’ inputs. Ensure that Storage_Policy, Network_Policy, Boot_Policy_LAN and Boot_Policy_SAN entries are correct.
- Once this is done , you can move to SAN zoning.
2. Configure SAN Zoning
- In this step, a new workflow task will be created in order to configure SAN zoning. Rather than use a specific user input for this task, output variables from the previous ‘Create UCS Service Profile’ workflow task will be used as input items for this task.
- Open the workflow and search for the workflow task ‘configure san zoning’. Drag the storage workflow task into the work area and map the following user inputs.
Create Flexible Volume
- Create a ‘Create Flexible Volume’ workflow task in order to build a NetApp flexible volume and provision it for the required size.
- Once again, reconfigure the workflow such that the success criteria from the ‘Configure SAN Zoning’ workflow task proceeds to ‘Create Flexible Volume’.
- Using the NetApp ONTAP ‘Create LUN’ workflow task, create a LUN located within the volume created during the previous step. In order to do this, map the Volume Name user input to the output from the previous ‘Create Flexible Volume’ workflow task.
- Next, enter the LUN details, ensuring that the configured size is less than that of the volume created in the previous step.
- As before, re-map the workflow designer flow so that the successful output of the Create Flexible Volume workflow task flows into this task.
3. Create Initiator Group
- Create a ‘Create Initiator Group’ workflow task in order to build a NetApp ONTAP Initiator Group. Map the ‘Filer Identity Name’ attribute to the OUTPUT_FILER_IDENTITY output variable as supplied by the ‘Create NetApp Flexible Volume’ workflow task as created earlier in this workflow.
- Once again, modify the the workflow designer flow so that the successful output from ‘Create LUN’ flows into ‘Create Initiator Group’
4. Add Initiator to Initiator Group
- Create an ‘Add Initiator to Initiator Group’ workflow task and map the ‘Initiator Group Name’ entry to the OUTPUT_IGROUP_IDENTITY output variable from the ‘Create NetApp Initiator Group’ workflow task and map the ‘Initiator Name’ entry to the SP_VHBA1 output variable from the ‘Create UCS Service Profile’ workflow task created earlier in this flow.
- Repeat this task for in order to add initiator name entry for SP_VHBA2. Once done, re-map the successful output from the two Create Initiator Group workflow tasks so that they flow into each other as follows:
- Move onto the next step in order to map the created LUN to the initiator group.
And the rest of steps are fairly similar to above. Create a Workflow and map the User Input Mappings".
5. Map LUN to Initiator Group
6. Modify UCS Boot Policy LUN ID
7. Select UCS Server
8. Associate UCS Service Profile
9. Power On UCS Server
10. Modify UCS Service Profile Boot Policy
11. Add VLAN to Service Policy
12. Disassociate UCS Service Profile
13. Wait for Specified Duration
14. Associate UCS Service Profile
15. Power On UCS Server
16. Register Host with VCenter
I'm sure this would be extremely helpful for our deployment when we come to this step. Unfortunately we are still working bit by bit to gain control over Server and SAN systems as they are currently managed by desgnated teams. That's why our very intial step is to showcase the automation and orchestration in network layer and gain consensus with other teams.
Again thanks a million!
Could you advise on how to set-up the following?
3. Periodic reports
4. chardback currency
5. VM locking security
Looking forward to your answer.
Thanks for the participation in this Ask the Expoert session.
I am not too sure if you are looking for a specific "Notification" task but generally speaking Nofitication can be configured at various diffenent task and stage. One example I can share with you here is when we configure vDC, we can define virtual resources, operational details, rules, and policies to manage specific group requirements. In here you can also define approvers specific to a vDC. The approver for a particular vDC must approve the request from users for VM provisioning and this task can be notified during the provisioning process by providing email contact detail. Therefore, configuring "Mail Setup" Under "System Administration" is the pre-requisite for the Notification.
Go to Administration > System Administration > Mail Setup tab
VDC creation and Approvers and Contacts
3. Periodic Report
You can view the report from various different places. First place you would find is Virtual menu - Virtual, Storage, Network, and VDI. In here Admin can set automatic or manual refresh for report widget data. For automatic refresh admin has an option to customize the automatic refresh time interval (5 mins to 60 mins).
If you look for the Report Customization, you can configure this from select Administration > System Administration > Reports Customerization tab :
4. Chargeback Currency :
- If you need to understand how to configure chargeback, please refer to below config guide
However, if you only need to know how to set the chargeback currency, then go to system parameter tab below
Administration > System Adminsitration > System Parameter tab
5. VM Locking Security :
This can be configured from the same place as above.
BTW, I skipped Alert question. I'm not really clear what you want to know about the Alert in UCS director. If you can be a little bit more specific, I will assist you further.
Thanks for the participation in this Ask the Expoert session.
Yes, Service End User can add Label to VM at User service portal.
Here is how to :
Assuming you already have VM provisioning policy,
- Go to Policies in UCSD Director main GUI and select VM provisioning policy under VMware System Policy.
- highlight the policy and click "Edit". In here you will see "End User VM Name or VM Prefix".
- Tick "End User VM Name or VM Prefix" and click save.
- Next login into User Self Service portal
- Now you will see the VM Name or VM Prefix field to enter during the Deployment Configuraiton stage as below.
I hope this helps you a little.
Thank you for your reply. After deploy VM user cannot change this value.
Is this correct ?
Another question.I find VM Label field at user portal. Can user modify this value ?
I know admin can modify this value from admin portal.
I don't think you can change that value after the reployment.
For the VM Label you can modify from admin portal, not from user portal.
hey Andrew, thanks for your help. One more thing please. Could you also advise how I can check on the general status of service or database? thanks very much!
No problem. You can view the service and DB status from Admin GUI as below :
Go to "Administration" > "System Administration" and then look for "Support Information" tab and then click "System Information and Logs Link Page" and it will bring you the "System Information and Logs Link Page" as below.
Click either "System administration (Basic) or (Advanced). Then you will see service and DB status as below :
If you need to perform DB backup or restore task, you can do it from shelladmin access as below :
ssh to UCS director with shelladmin account and then you will see something similar to below shell menu
Menu 2 is to display service status and 3 and 4 are stop and start the services.
Menu 5, 6, 7, and 8 are related to Database.
General rule of thumb for DB is make sure to have a good DB backup before any upgrade.
I hope this helps you a little bit more.