api changes from R80.40 to R81.10 batch files

[ERTK]

hello,

Some month ago we deploy a new cluster on R80.40 (SMS and GW) by api and csv batch file (it had 40 interfaces).

I am trying to make the same procedure, with the same csv file (with another name, ips etc) in a SMS R81.10 jumbo 45 and it gives me an error:

the cluster network xxx is missing a valid ip address for some of the members.

I run the old SMS (R80.40) run the same commands and csv file...and it works. 

I review the api from R80.40 and R81.10 and it seems that the commands for adding interfaces on simple cluster are identical.....any kind of suggestion will be appreciated.

[PhoneBoy]

What is the precise command line you're using and what does the CSV file look like?
It could very well be a bug and thus a TAC case may be needed.

[ERTK]

i create cluster 

and i copy paste example from api v1.8:

mgmt_cli -r true set simple-cluster name "fw-vip" interfaces.add.name "eth3" interfaces.add.ip-address "10.10.10.1" interfaces.add.ipv4-mask-length "24" interfaces.add.interface-type "cluster" interfaces.add.topology "INTERNAL" interfaces.add.anti-spoofing "true" members.update.1.name "fw-i" members.update.1.interfaces.name "eth3" members.update.1.interfaces.ipv4-address "10.10.10.2" members.update.1.interfaces.ipv4-network-mask "255.255.255.0" members.update.2.name "fw-ii" members.update.2.interfaces.name "eth3" members.update.2.interfaces.ipv4-address "10.10.10.3" members.update.2.interfaces.ipv4-network-mask "255.255.255.0" --format json

message:

{
"tasks" : [ {
"task-id" : "01234567-89ab-cdef-add6-8f7057328900",
"task-name" : "set simple-cluster",
"status" : "failed",
"progress-percentage" : 100,
"suppressed" : false,
"task-details" : [ {
"fault-message" : "Action Failed due to an Internal Error"
} ]
} ]

[Tal_Paz-Fridman]

Hi

The command you're using is set and not add meaning the Cluster already existed?

If so what command did you use to create it?

[ERTK]

it is correct, i had created before cluster with command:

mgmt_cli add simple-cluster name "fw-vip" ......

that command was executed sucessfull

[Tal_Paz-Fridman]

Thank you. Could you please share the full command used to create it? I would like to run the full flow.

Thanks

[ERTK]

mgmt_cli -r true add simple-cluster name "fw-vip" version "R80.40" ip-address "10.10.11.251" os-name "Gaia" cluster-mode "cluster-xl-ha" firewall true interfaces.1.name "eth2" interfaces.1.ip-address "10.10.11.251" interfaces.1.network-mask "255.255.255.0" interfaces.1.interface-type "cluster" interfaces.2.name "Sync" interfaces.2.interface-type "sync" interfaces.2.topology "INTERNAL" interfaces.2.topology-settings.ip-address-behind-this-interface "network defined by the interface ip and net mask" interfaces.2.topology-settings.interface-leads-to-dmz false members.1.name "fw-i" members.1.one-time-password "vpn123" members.1.ip-address "10.10.11.2" members.1.interfaces.1.name "eth2" members.1.interfaces.1.ip-address "10.10.11.2" members.1.interfaces.1.network-mask "255.255.255.0" members.1.interfaces.2.name "Sync" members.1.interfaces.2.ip-address "1.1.1.1" members.1.interfaces.2.network-mask "255.255.255.248" members.2.name "fw_ii" members.2.one-time-password "vpn123" members.2.ip-address "10.10.11.3" members.2.interfaces.1.name "eth2" members.2.interfaces.1.ip-address "10.10.11.3" members.2.interfaces.1.network-mask "255.255.255.0" members.2.interfaces.2.name "Sync" members.2.interfaces.2.ip-address "1.1.1.2" members.2.interfaces.2.network-mask "255.255.255.248"

[Tal_Paz-Fridman]

In the add command one object is fw-i and the other is fw_ii  --- hyphen vs. underscore.

later in the set command the object name is shown with hyphen fw-ii while in the add command it is shown with underscore fw_ii

Try running:

mgmt_cli -r true set simple-cluster name "fw-vip" interfaces.add.name "eth3" interfaces.add.ip-address "10.10.10.1" interfaces.add.ipv4-mask-length "24" interfaces.add.interface-type "cluster" interfaces.add.topology "INTERNAL" interfaces.add.anti-spoofing "true" members.update.1.name "fw-i" members.update.1.interfaces.name "eth3" members.update.1.interfaces.ipv4-address "10.10.10.2" members.update.1.interfaces.ipv4-network-mask "255.255.255.0" members.update.2.name "fw_ii" members.update.2.interfaces.name "eth3" members.update.2.interfaces.ipv4-address "10.10.10.3" members.update.2.interfaces.ipv4-network-mask "255.255.255.0" --format json

[ERTK]

hello, thanks for your answer...this is a copy paste problem mine. i did it with the same name on both commands 

[ERTK]

here the error that report on smartconsole:" 

Task: set simple-cluster
Initiator: WEB_API
Start Time: 26/04/2022 14:59
Completed: 26/04/2022 14:59
Status: Failed
--------------------------------------------------------------------------------
The cluster network eth3 is missing a valid IP address for some of the members.
--------------------------------------------------------------------------------

"

[Tal_Paz-Fridman]

Both commands worked for me (after fixing the underscore - hyphen issue)

[ERTK]

in which SMS version and jumbo?

[Raj_Khatri]

ERTK, did you get this working?  Facing a similar issue with APIs working in R80.40 no longer work on R81.10.

[ERTK]

hi, unfortunately issues with this command "mgmt_cli add simple-cluster name"  couldn't solve it, i openned a case with TAC but it didnt solve .

we solved the situation using another approach with other API command , First we create a simple cluster with basics intefaces and then we edit this object with "mgmt_cli set simple-cluster" and bach file.

This approach work for us.

[Amir_Senn]

Hi,

Did you try changing the underscore/hyphen?

I think there were no relevant changes in the API but maybe some validation was added in another place that might cause this.

[ERTK]

It was a transcription error. i changed some details of command in order to hide customer information.

I used the same command and csv in one SMS R80.40 and R80.10.....the first one worked and second failt. 

[Amir_Senn]

It looks like this is unrelated to the API itself and some new validation were added in R81.10 that caused this issue to happen.

[Amir_Senn]

Hi Raj,

Can you also share the full command?

[Raj_Khatri]

1. Get the session ID

curl -k --request POST 'https://CMAIP:443/web_api/login' --header 'Content-Type: application/json' --data-raw '{"user" : "USERNAME", "password" : "PASSWORD", "domain" : "DOMAIN"}'

2. Use session ID from above to get routes

curl -k --request POST 'https://CMAIP:443/web_api/v1/run-script' --header 'X-chkp-sid: SESSIONID' --header 'Content-Type: application/json' --data-raw '{"script-name": "Script Example: Show Routes", "script": "netstat -rnv", "targets": ["R80-FW1"]}'

On non-SMB firewalls, this succeeds and returns a task ID.  However, on SMB firewalls (1100, 1400, 1500) it fails with a 'generic_err_object_not_found' error.

[Hugo_vd_Kooij]

Can you run R80 on 1100 appliances? Not very likely. So how can they even be part of the groep R80-FW1 ?

You can try to run scripts from Smart Console first. If that fails you propably get an error message telling why it can't do it.

But as a general hint: SMB devices are not that much into R80 and thus not that much into automation.

[PhoneBoy]

It may be the command needs to be with full path to execute successfully on SMB appliances.