API service wont start

[Itzko]

Hello

I'm using R80.40 on VMware

After fresh install I'm able to connect the web UI just fine and deploy updates.

The problem starts when I'm trying to connect via smart console, I'm getting the "Unable to connect to server.
Please make sure that all processes of the server are up and running." error.

After short investigation I figured the API service isn't starting properly here is the api status output:

FW1> api status
cat: /web/conf/extra/httpd2-webapi.conf: No such file or directory

API Settings:
---------------------
Accessibility: Not set
Automatic Start: Unknown

Processes:

Name State PID More Information
-------------------------------------------------
API Stopped
CPM Started 18297 Check Point Security Management Server is running and ready
FWM Stopped
APACHE Started 17271

Port Details:
-------------------
JETTY Internal Port: 10490
APACHE Gaia Port: 443
Apache port retrieved from: httpd-ssl.conf

Profile:
------------
Machine profile: Standalone 4GB to 7GB
CPM heap size:
API heap size:

--------------------------------------------
Overall API Status: The API Server Is Not Running!
--------------------------------------------

Notes:
------------
To collect troubleshooting data, please run 'api status -s <comment>'

When im trying the api start command im getting:

FW1> api start
2021-Mar-13 21:41:40 - Starting API...
2021-Mar-13 21:41:40 - WARNING: File mgmt_api_profile_settings.xml not found.
2021-Mar-13 21:41:40 - WARNING: Waiting for mgmt_api_profile_settings.xml to be created.

And nothing happens

Any help? 

[PhoneBoy · Accepted solution]

It’s likely you have not allocated enough RAM to the VM.
For a standalone gateway (management and gateway on same VM), I would not allocate any less than 16GB.
It may work with less, but if you don’t allocate enough RAM, the API server will either not start or be slow. 

[Vincent_Bacher]

I often buil managagements in lab on vm and when api did not start, reason was always because I forgot to set RAM to 16 GB.

Absolute minimum is 8 but here I faced either services needed ages to start or api did not start at all.

[Itzko]

Thank you for your answer.

I didn't mention this before but the machine is part of a lab and running on 4GB of RAM.

I did deployed successfully previous machines for labs with only 4 GB of RAM with no problems at all.

I re-install the problematic machine with only 4 GB again and got it to work,

The main two differences were that I didn't deployed the jumbo update before trying to log via smart console

and I changed the time zone on the windows machine that the smart console installed on to match the time zone of the R80.40 machine.

I don't know if one of those changes made the difference but its working now with only 4 GB of RAM.

Thanks again!

[PhoneBoy]

In R77.30 and earlier, 4GB was sufficient for a standalone gateway.
From R80, the minimum required RAM for a standalone gateway is 8GB--this is documented in the Release Notes.
That said, 16GB is highly recommended for a standalone gateway.
With management on a separate VM (8GB+ RAM), 4GB of RAM for a gateway VM may be sufficient.

[the_rock]

Guys that responded are 100% correct, but just curious, what does it show when doing cpwd_admin list on the mgmt server?

Andy