Ошибка could not create socket

Hello,

We are experiencing occasional 421 Could not create socket notices. This is causing some of our ftp jobs to fail and need to be resent. We are a low volume user, around 40 logins daily transferring 5 files. The files vary in size from 500K to 15-20 meg. The connections are spaced out through the day so they don’t all hit at once. We are using the current version 0.9.60.2. We are using passive mode without specifying a port range since everything is internal, no external connections to the internet, everything is open. Here is a sample of what we are seeing. I have changed the login and ip address. When the PASV command is issued is the could not create socket coming from my FTP server or the client? Should I define a port range for Filezilla to use?

(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> 230 Logged on
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> CWD
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> 250 CWD successful. «/» is current directory.
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> TYPE I
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> 200 Type set to I
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> PASV
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> 421 Could not create socket.
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> QUIT
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> 221 Goodbye
(004272) 9/21/2018 8:45:21 AM — anyftp (11.200.10.245)> disconnected.

Thanks

Rob

We are using passive mode without specifying a port range since everything is internal, no external connections to the internet, everything is open.

Consider limiting the Passive range to 49152-65535, even if it is internal. The reason is that only the mentioned range is most probably unused (meant for temporary, ephemeral usage). All lower ports might at least be occupied partially by other services. When FileZilla tries to create a passive socket on such an occupied or blocked port, the result is the error 421 as you experienced.

Things to consider:
— Antivirus programs or other security software can block sockets. Sockets can also be occupied by other services.
— The error shown starts with a «4», thus it is of temporary nature. In such situations, your scripts should simply retry the transfer. Aborting the scrips is only correct in case of errors starting with «5» (permanent error).

When the PASV command is issued is the could not create socket coming from my FTP server or the client?

Lines starting with a response code always come from the server.

Btw. you don’t need to obfuscate internal IP addresses (starting with 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16). These are not unique in any way and cannot be used for identification.

Consider limiting the Passive range to 49152-65535, even if it is internal. The reason is that only the mentioned range is most probably unused (meant for temporary, ephemeral usage). All lower ports might at least be occupied partially by other services. When FileZilla tries to create a passive socket on such an occupied or blocked port, the result is the error 421 as you experienced.

That’s no longer necessary since version 0.9.51, since then if no range has been manually been configured, 49152 through 65535 is used.
FileZilla Server allocates ports in a manner that automatically prevents reuse of ports within the mandatory TIME_WAIT interval (see TCP RFCs). In case a port is selected that is already used by a program other than FileZilla Server, it retries up to 15 times to find an alternative before giving up.

I was able to get Symantec turned off and I am still getting random 421 Could not create socket errors. Next thing to try is putting in a custom port range of 52,000 to 53,000. My other thought is to go to a slightly older release. Any other suggestions?

Thanks

Rob

Hi,

I’m going to close the loop on this issue. I disabled Symantec and I still received the 421 socket error. I then set the passive range to 52,000 — 53,000. No 421 socket errors for almost 2 weeks.

Rob

I’m installing through brew.

The error message(s):

When I call

postgres

I receive the error

LOG:  could not bind IPv6 socket: Address already in use
HINT:  Is another postmaster already running on port 5432? If not, wait a few seconds and retry.
LOG:  could not bind IPv4 socket: Address already in use
HINT:  Is another postmaster already running on port 5432? If not, wait a few seconds and  retry.
LOG:  could not bind IPv6 socket: Address already in use
HINT:  Is another postmaster already running on port 5432? If not, wait a few seconds and retry.
WARNING:  could not create listen socket for "localhost"
FATAL:  could not create any TCP/IP sockets

Running

postgres -D /usr/local/var/postgres

Creates the same error

createdb

prompts me for my password twice, then says,

createdb: could not connect to database template1: FATAL:  password authentication failed for user "thomasmurphy"

What I’ve tried
* Uninstalling and reinstalling the pg gem, uninstalling and reinstalling postgres
* Updating command line tools
* Satisying all of brew doctor’s needs
* editing pg_hba.conf’s permissions
* editing .bash_profile to be sure it includes the above directories

I have Postgres running perfectly on my work computer, also on Mavericks, so this strikes me as somewhat of an edge case. What other vectors can I poke at to try to solve this?

Linux: Make sure that your loopback interface is up by running the ifconfig at the command line. If it’s not up, you can use /sbin/ifconfig lo 127.0.0.1 to bring it up.

A machine reboot resolved the error for me.

There are multiple scenarios and places to check in this case. Got lost once because of this.
A usual default hosts file should look like one below:

[root@localmachine ~]# cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6

To confirm localhost is able to resolve, test using ping:

[root@localmachine ~]# ping localhost

Check cat /etc/nsswitch.conf|grep hosts and confirm files is there:

#     hosts: files dns
#     hosts: files dns  # from user file
#hosts:     db files nisplus nis dns
hosts:     files mdns4_minimal [NOTFOUND=return] dns myhostname

Check permissions of /etc/hosts file, it should be 644 ( readable for all processes/users` ):

[root@localmachine ~]# ls -al /etc/hosts
-rw-r--r--. 1 root root 240 Aug 14 08:04 /etc/hosts

For those who have enough time, I’ve created a detailed blog post here:
https://coffeewithbytes.com/blog/postgres-hostsfile-permissions/

I’ve mentioned how I checked and came to a conclusion.

answered Aug 16, 2019 at 2:58

Did you installed another version of Postgres previously? In my case, I was trying to execute:

sudo systemctl restart postgresql

that was incorrect!! finally I found the correct one is:

sudo systemctl restart postgresql-12

I had two versions of Postgres…

answered Jul 18, 2020 at 13:21

wireguard1: Could not create IPv4 socket

Tue Mar 22, 2022 8:15 pm

Re: wireguard1: Could not create IPv4 socket

Tue Mar 22, 2022 8:21 pm

Are you using a port which is already been used by another service ?

Re: wireguard1: Could not create IPv4 socket

Tue Mar 22, 2022 8:29 pm

Re: wireguard1: Could not create IPv4 socket

Tue Mar 22, 2022 8:33 pm