cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
39
Views
0
Helpful
1
Replies
Highlighted
Cisco Employee

NED errors and fastmap

 

Hi Team,

 

We see that when we get ned_prepare_cli errors  either during service activation or service rollback. That it leaves the the services defined in the CDB database, but it does print us a nice message. With chained services then we have lots of clean up of the CDB to get ready to retest.

 

Is there a recommended way to capture these errors to try to leave the system in a better state?

 

In an ncs system-install – is there a recommended way to reset the database? To assit with clean up.

 

Sometimes these errors do not manifest themselves until we are applying or rolling back these chained services. When we test the services independently we do not have an issue. So they are complexity driven. In our services we have to apply configuration to our STAROS device in multiple services in order to activate the licensing correctly – before configuring the devices. So we use reactive fastmap for this.

 

Thanks in advance, JA

 

1 ACCEPTED SOLUTION

Accepted Solutions
Highlighted
Cisco Employee

 

On 23/01/16 02:25, Julie Ann Connary (jconnary) wrote:

 

> Hi Team,

 

>

 

> We see that when we get ned_prepare_cli errors  either during service

 

> activation or service rollback. That it leaves the the services

 

> defined in the CDB database, but it does print us a nice message. With

 

> chained services then we have lots of clean up of the CDB to get ready to retest.

 

>

 

 

Sounds as if you have bugs in the NED, either in YANG, Java or both.

 

 

> Is there a recommended way to capture these errors to try to leave the

 

> system in a better state?

 

>

 

> In an ncs system-install – is there a recommended way to reset the

 

> database? To assit with clean up.

 

 

ncs-backup is a fast and convenient way of restoring the system back to good known state.

 

 

$ man ncs-backup

 

 

 

>

 

> Sometimes these errors do not manifest themselves until we are

 

> applying or rolling back these chained services. When we test the

 

> services independently we do not have an issue. So they are complexity

 

> driven. In our services we have to apply configuration to our STAROS

 

> device in multiple services in order to activate the licensing

 

> correctly – before configuring the devices. So we use reactive fastmap for this.

 

>

 

 

Ok, I see, we had similar issues with bugs i the ASA NED and applying config in stages during CVPN development.

 

 

It's especially important to make revert work properly in the NED.

 

You'll have good CLI output from the device in the ncs java log, it should be clean.

 

 

/klacke

 

View solution in original post

1 REPLY 1
Highlighted
Cisco Employee

 

On 23/01/16 02:25, Julie Ann Connary (jconnary) wrote:

 

> Hi Team,

 

>

 

> We see that when we get ned_prepare_cli errors  either during service

 

> activation or service rollback. That it leaves the the services

 

> defined in the CDB database, but it does print us a nice message. With

 

> chained services then we have lots of clean up of the CDB to get ready to retest.

 

>

 

 

Sounds as if you have bugs in the NED, either in YANG, Java or both.

 

 

> Is there a recommended way to capture these errors to try to leave the

 

> system in a better state?

 

>

 

> In an ncs system-install – is there a recommended way to reset the

 

> database? To assit with clean up.

 

 

ncs-backup is a fast and convenient way of restoring the system back to good known state.

 

 

$ man ncs-backup

 

 

 

>

 

> Sometimes these errors do not manifest themselves until we are

 

> applying or rolling back these chained services. When we test the

 

> services independently we do not have an issue. So they are complexity

 

> driven. In our services we have to apply configuration to our STAROS

 

> device in multiple services in order to activate the licensing

 

> correctly – before configuring the devices. So we use reactive fastmap for this.

 

>

 

 

Ok, I see, we had similar issues with bugs i the ASA NED and applying config in stages during CVPN development.

 

 

It's especially important to make revert work properly in the NED.

 

You'll have good CLI output from the device in the ncs java log, it should be clean.

 

 

/klacke

 

View solution in original post

Content for Community-Ad
Cisco Community August2020 Spotlight Award Winners
This widget could not be displayed.