cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
2037
Views
0
Helpful
2
Replies

"transport timeout; closing session" when trying to deploy configuration via NSO

jamie_quigg
Level 1
Level 1

Hi,  We are getting a weird error when trying to commit a configuration via NSO: "transport timeout; closing session"  Does anyone know what the issue/resolution is for this?

1 Accepted Solution

Accepted Solutions

alam.bilal
Cisco Employee
Cisco Employee

Hi Jamie,

As mentioned in the email response earlier:

The first place to start is to look at the log files. All logs are placed in the directory as configured in “ncs.conf”. Usually these are the files to start with:

ncs-java-vm.log

  1. ncs.log
  2. devel.log
  3. ncserr.log

I’m guessing the issue is timeout while accessing a device? This can happen with slow devices and/or with very large configs. First ensure NSO can access the device[s].

admin@ncs(config)# devices device alu0 disconnect

admin@ncs(config)# devices device alu0 connect

The timeouts when accessing the device (connection, read, write) can be adjusted globally or per device:

admin@ncs(config)# devices global-settings read-timeout

admin@ncs(config)# devices global-settings write-timeout

admin@ncs(config)# devices global-settings connect-timeout

For per-device settings:

admin@ncs(config)# devices device alu0 read-timeout


Other things to consider, is this for a particular device? Or is this only for a particular configuration sequence? If the latter then it could be a NED issue. Then we collect NED traces and raise a case:

admin@ncs(config)# devices device ios0 trace raw

admin@ncs(config-device-ios0)# disconnect

admin@ncs(config-device-ios0)# connect

result true

info (admin) Connected to ios0 - 127.0.0.1:10022

A new trace file for the device will be created in the logs directory.

View solution in original post

2 Replies 2

alam.bilal
Cisco Employee
Cisco Employee

Hi Jamie,

As mentioned in the email response earlier:

The first place to start is to look at the log files. All logs are placed in the directory as configured in “ncs.conf”. Usually these are the files to start with:

ncs-java-vm.log

  1. ncs.log
  2. devel.log
  3. ncserr.log

I’m guessing the issue is timeout while accessing a device? This can happen with slow devices and/or with very large configs. First ensure NSO can access the device[s].

admin@ncs(config)# devices device alu0 disconnect

admin@ncs(config)# devices device alu0 connect

The timeouts when accessing the device (connection, read, write) can be adjusted globally or per device:

admin@ncs(config)# devices global-settings read-timeout

admin@ncs(config)# devices global-settings write-timeout

admin@ncs(config)# devices global-settings connect-timeout

For per-device settings:

admin@ncs(config)# devices device alu0 read-timeout


Other things to consider, is this for a particular device? Or is this only for a particular configuration sequence? If the latter then it could be a NED issue. Then we collect NED traces and raise a case:

admin@ncs(config)# devices device ios0 trace raw

admin@ncs(config-device-ios0)# disconnect

admin@ncs(config-device-ios0)# connect

result true

info (admin) Connected to ios0 - 127.0.0.1:10022

A new trace file for the device will be created in the logs directory.

I had an issue with the "transport timeout; closing session" when migrating services from monolithic NSO to LSA NSO on netsim devices.

I solved the issue by increasing those 3 time-out settings in both CFS and RFS NSO.

admin@ncs(config)# devices global-settings read-timeout 300

admin@ncs(config)# devices global-settings write-timeout 300

admin@ncs(config)# devices global-settings connect-timeout 300

 

Thanks for the help.