Process fastrpc.cgi taking all CPU when triggering certain actions in Virtualmin

SYSTEM INFORMATION
OS type and version Ubuntu Linux 16.04.7
Webmin version 1.999
Virtualmin version 7.7-3
Processor information Intel(R) Xeon(R) CPU E5-2686 v4 @ 2.30GHz, 2 cores
Real memory 3.85 GiB total

Note:
I’m using a custom port to access webmin. This server has a few years and never had this problem.

Process running command fastrpc.cgi is taking all available CPU and makes the server unusable. This happened when the following actions were triggered:

1. When removing a virtualmin server that is already disabled:
In the two following steps the process hanged and I had to kill process fastrpc.cgi in order for it to continue. I got the following error messages:

Removing SSL from Nginx virtual host …
… Nginx SSL website failed! : virtualmin-nginx-ssl::feature_delete failed : Error reading response length from fastrpc.cgi :

Deleting DNS zone …
… DNS domain failed! : Error reading response length from fastrpc.cgi :

2. When accessing Server Configuration > DNS Records page

The page hangs with message “can not stablish connection to server”.

or sometimes the error is:

“Error reading response length from fastrpc.cgi”

3. When creating a new virtualmin server

The page hangs with message “can not stablish connection to server”.


For all three of those previous actions the process running fastrpc.cgi takes all CPU memory until the usage registered of the CPU is 99%.
I thought it was a problem with no enough memory but when I upgraded the server and added more RAM/cores it also crashes in the same actions.

Additional note:

I did run “Re-check Configuration” and got: … your system is ready for use by Virtualmin

Could it be related to DNS and a possible timeout trying to resolve something there?

News:

Ports 10000 to 10100 were opened but still not able to fix this issue.

4G of real memory isn’t a lot. How much swap space do you have?

Barely enough to run nginx let alone the rest of the basic system processes. Didn’t Virtualmin cough :scream: during install

I run on 4Gig no problem(apache). Maybe OP has the virus scanner running, that can take up to 2gig ram.
I have 16 virtual servers.

does nginx need more ram? VM should easily install on 1G.

Ok so I increased my swap as you can see here:

$ free -h
              total        used        free      shared  buff/cache   available
Mem:           3,9G        172M        3,5G         31M        204M        3,4G
Swap:          9,0G        770M        8,2G

Then in console I ran:

$ virtualmin get-dns --domain mydomainname.com

And I let it run until it stopped with following error:

Error: Error reading response length from fastrpc.cgi : 
Error
-----
Error reading response length from fastrpc.cgi : 
-----

fastrpc-cgi again took all RAM, then it was taking all SWAP slowly and finally crashed.

Btw, this happens with any domain name I have hosted in this server. All small/personal sites.

You should be looking at using a later OS, this version is no longer supported.
Are you running any autobackup?
Google the issue there are other reports of this, have you read them?

ok found out what it does , might trigger something.

The fastrpc.cgi script is used for RPC calls between webmin
servers, for example when you use the Cluster category
modules. It uses ports 10001 and up - the maximum is
determined by how many instances of fastrpc.cgi run at once,
but shouldn’t be more than 10100.
So you should allow all of those ports on your firewall as
well …

So have got registered servers?
I see the fastrcp when you register one

  • No servers registered so that section loads normally.
  • I do have ports 10000 - 10100 open using firewalld.
  • Yeah I might choose to upgrade this OS if no other option works for me.
  • I’m not running any autobackup. Memory is pretty clear and only used when fastrpc.cgi runs.

In other news:

  • I see fastrpc.cgi also gets trigggered when I try to visit SSL certificates page and it crashes in SSL Let’s Encrypt tab as well. I wonder if it has to do anything with it.
  • It’s pretty late over here, I’ll have more updates tomorrow.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.