Maintainance mode
Steps
  • This is the main dashboard in the Apstra solution, notice there are no anomalies. For the purpose of this demo, let us perform some maintenance on a device.
  • There are two Spine devices, and three racks. Two of the racks have redundant switches in them, and the third rack has a single switch in it. Let us perform maintenance on spine two. To begin, click on the staged tab.
  • Select Spine 2
  • Note that this is the normal operational state for this device.
  • Select the drain mode
  • Click on save icon to confirm the change
  • The command is being queued into the uncommitted tab and ready to invoke this process.
  • Commit the change
  • Abstract aids us in keeping accurate documentation by allowing us to label the maintenance events.
  • The commit process results in route map adjustments that removes the device as a path and the routing table of all the other devices.
  • As this proceeds, the telemetry table begins to show the effects that are resulting from the mode change. BGP peering sessions to Spine to have dropped, and changes in the route tables of the other devices are being indicated. Spine two is shown as being in the Drain mode. The red anomalies are showing the expected changes and will soon clear as the fabric adjusts to the new state where spine two has been gracefully isolated and all traffic is traversing the other devices in the fabric.
  • An IBA probe closely watches the interface utilization on during devices. This provides visibility into the point in time when implied traffic has completed its journey.
  • Choose the Drain status from the drop down menu
  • Click on Analytics
  • Once the traffic level on Spine drops below a customizable threshold, the probe anomaly clears and indicates the spine two is ready to be serviced. This typically takes a few moments, depending on the level of utilization at the time the process is initiated.
  • Click on Dashboard
  • Click on active
  • Moving into the Active tab shows that spine two remains in this state, and we can now perform the required maintenance procedure on this switch
  • Now that maintenance is complete, spine two can be returned to operation. The system will return the route maps to their original state, and spine two's capacity will be returned to the fabric. Now we reverse the process we performed earlier, and we return spine two to the deployed state. 
  • Click to change the mode
  • Change the mode to Deploy
  • Click to save the selection
  • As earlier, the instructions for this device are queued. This is reviewed in the uncommitted tab. Pressing commit begins the process of bringing Spine two back into the fabric.
  • 26
  • Commit changes
  • Let us check the active tab
  • All the devices are back in the deploy mode
  • A quick review of our top-level dashboard shows that all indications are green, and drain config is not present on any devices.
0/3
There are a few loopholes and inconsistencies (for example {spoiler alert}: how the actual f@#k does the clan keep 3 giant ancient anacondas fed and alive ffs??) in conjunction with a few sub-par performances, but, overall its a pleasant spinoff and accompaniment to the other 2 films in the franchise, and does so with considerably less focus on firearms and more on the action one might expect to see from a strong culturally Japanese standpoint.
Prev
Next

Please enter details to continue

By submitting this form, I confirm that I have read and agree to the Privacy Statement.

Please enter your email to continue
In this Demo
  • %3Cp%3EThis%20is%20the%20main%20dashboard%20in%20the%20Apstra%20solution%2C%20notice%20there%20are%20no%20anomalies.%20For%20the%20purpose%20of%20this%20demo%2C%20let%20us%20perform%20some%20maintenance%20on%20a%20device.%3C/p%3E
  • %3Cp%3EThere%20are%20two%20Spine%20devices%2C%20and%20three%20racks.%20Two%20of%20the%20racks%20have%20redundant%20switches%20in%20them%2C%20and%20the%20third%20rack%20has%20a%20single%20switch%20in%20it.%20Let%20us%20perform%20maintenance%20on%20spine%20two.%3C/p%3E%0A%3Cp%3ETo%20begin%2C%20click%20on%20the%20staged%20tab.%3C/p%3E
  • Select Spine 2
  • %3Cp%3ENote%20that%20this%20is%20the%20normal%20operational%20state%20for%20this%20device.%3C/p%3E
  • Select the drain mode
  • Click on save icon to confirm the change
  • %3Cp%3EThe%20command%20is%20being%20queued%20into%20the%20uncommitted%20tab%20and%20ready%20to%20invoke%20this%20process.%3C/p%3E
  • Commit the change
  • %3Cp%3EAbstract%20aids%20us%20in%20keeping%20accurate%20documentation%20by%20allowing%20us%20to%20label%20the%20maintenance%20events.%3C/p%3E
  • %3Cp%3EThe%20commit%20process%20results%20in%20route%20map%20adjustments%20that%20removes%20the%20device%20as%20a%20path%20and%20the%20routing%20table%20of%20all%20the%20other%20devices.%3C/p%3E
  • %3Cp%3EAs%20this%20proceeds%2C%20the%20telemetry%20table%20begins%20to%20show%20the%20effects%20that%20are%20resulting%20from%20the%20mode%20change.%20BGP%20peering%20sessions%20to%20Spine%20to%20have%20dropped%2C%20and%20changes%20in%20the%20route%20tables%20of%20the%20other%20devices%20are%20being%20indicated.%20Spine%20two%20is%20shown%20as%20being%20in%20the%20Drain%20mode.%20The%20red%20anomalies%20are%20showing%20the%20expected%20changes%20and%20will%20soon%20clear%20as%20the%20fabric%20adjusts%20to%20the%20new%20state%20where%20spine%20two%20has%20been%20gracefully%20isolated%20and%20all%20traffic%20is%20traversing%20the%20other%20devices%20in%20the%20fabric.%3C/p%3E
  • %3Cp%3EAn%20IBA%20probe%20closely%20watches%20the%20interface%20utilization%20on%20during%20devices.%20This%20provides%20visibility%20into%20the%20point%20in%20time%20when%20implied%20traffic%20has%20completed%20its%20journey.%3C/p%3E
  • Choose%20the%20Drain%20status%20from%20the%20drop%20down%20menu
  • Click on Analytics
  • %3Cp%3EOnce%20the%20traffic%20level%20on%20Spine%20drops%20below%20a%20customizable%20threshold%2C%20the%20probe%20anomaly%20clears%20and%20indicates%20the%20spine%20two%20is%20ready%20to%20be%20serviced.%20This%20typically%20takes%20a%20few%20moments%2C%20depending%20on%20the%20level%20of%20utilization%20at%20the%20time%20the%20process%20is%20initiated.%3C/p%3E
  • Click on Dashboard
  • Click on active
  • %3Cp%3EMoving%20into%20the%20Active%20tab%20shows%20that%20spine%20two%20remains%20in%20this%20state%2C%20and%20we%20can%20now%20perform%20the%20required%20maintenance%20procedure%20on%20this%20switch%3C/p%3E
  • %3Cp%3ENow%20that%20maintenance%20is%20complete%2C%20spine%20two%20can%20be%20returned%20to%20operation.%20The%20system%20will%20return%20the%20route%20maps%20to%20their%20original%20state%2C%20and%20spine%20two%5C%27s%20capacity%20will%20be%20returned%20to%20the%20fabric.%20Now%20we%20reverse%20the%20process%20we%20performed%20earlier%2C%20and%20we%20return%20spine%20two%20to%20the%20deployed%20state.%C2%26nbsp%3B%3C/p%3E
  • Click to change the mode
  • Change the mode to Deploy
  • Click to save the selection
  • %3Cp%3EAs%20earlier%2C%20the%20instructions%20for%20this%20device%20are%20queued.%20This%20is%20reviewed%20in%20the%20uncommitted%20tab.%20Pressing%20commit%20begins%20the%20process%20of%20bringing%20Spine%20two%20back%20into%20the%20fabric.%3C/p%3E
  • 26
  • Commit changes
  • Let us check the active tab
  • All the devices are back in the deploy mode
  • A%20quick%20review%20of%20our%20top-level%20dashboard%20shows%20that%20all%20indications%20are%20green%2C%20and%20drain%20config%20is%20not%20present%20on%20any%20devices.