sorin's questions

Assuming that reinstalling from zero is not practical, how do I perform an update from CentOS 7 to CentOS 8?

I tried to do a yum install http://mirror.bytemark.co.uk/centos/8/BaseOS/x86_64/os/Packages/centos-release-8.0-0.1905.0.9.el8.x86_64.rpm but apparently that does not work as one would expect. Running yum update after this puts yum into an endless recursive attempt to resolve dependencies.

We do have a Google Apps domain and we want to configure a Radius server in order to allow people to login to our corporate WiFi or VPN using their corporate credentials (Google Apps).

To make the issue even more complex, we already have 2FA enabled for our Google Accounts. Still, if that would be a blocker we could ask people to use application-passwords for this purpose in order to avoid the 2FA issue.

So far I tried to use https://github.com/layeh/google-apps-radius and it did pass the radtest but it failed to work with our UniFi AP. While I contacted the vendor for details I am looking for alternatives in case this Radius server cannot really work. We are also considering hosted solutions.

What are the practical risks of enabling the unsecure DNS updates on Windows?

As far as I found enabling the unsecure DNS updates is a requirement for enabling DHCP Linux clients from registering their names with a FQDN.

I do want to know that are the practical risks involved in this in order to evaluate if that's ok to have these enabled or not.

As far as I know a machine would not be able to takeover another reserved name which would be the only real concern that I now.

Obviously it would be the DDOS but considering that we are talking about intranet here, I doubt this could be a real risk.

Do you have it enable on your domain or not? Did you ever had to disable it due to having some problems with it?

Where I work there is a practice of adding a t_ prefix to all temporary employees logins (sAMAccountName).

When someone switches from temp to perm or the the other way the prefix is add/removed.

While is ok with Windows and AD, this practice is messing most (if not all) other systems that are relying on AD/LDAP for authentication, as they are not able to detect a renamed account.

Is there any strong reason for having such a practice, aren't other alternatives that would play better with other systems?

I want to reconfigure a Synology running latest DSM 5 to work on normal HTTP ports.

The default configuration is to use port 5000 and 5001 for HTTPS and I do not like these because they are quite often blocked by most firewalls (in addition to the accessibility and visual aspects)

I am looking for a ssh / tmux solution that would act like this:

• if there is no session, create one
• if there is a session and nobody is connected to it, create another one

Mainly I want to be able to create new sessions to the same server, obviously if there is more than one session that has nobody connected to it, it should pick the first one.

This should enable me to put this as default command for ssh connections.

My current solution ssh -t 'tmux a || tmux || /bin/bash' doesn't work as expected because when you try to connect again it will connect to the existing session, and in this case I want a new one.

Using Cygwin environment on Windows is a very common solution for automation.

I am looking for a script that would automate the cygwin installation and enablement of ssh server.

I could try to write one but one of the problems is that I don't know what I can use to download line for downloading cygwin.

I have an nginx acting as a proxy in front of of a backend and I want to separate the requests made to a specific location to a separate file.

Still the requests should go to the same backend server but I don't want to see them inside the main access log.

Also I don't want to specify all the proxy_ stuff twice.

server {
    ...
    access_log  /var/log/nginx/jira.access.log full;
    error_log  /var/log/nginx/jira.error.log;
    client_max_body_size 150m;        

    location /special/ {
        }

    location / {

            # many lines of config params for proxy_...
            proxy_pass   http://dowa-02.example.com:8080;
            ...
    }
}

I do have an nginx server which responds to several domains and I do want to redirect all request to the main domain.

Example: website responding for xxx xxx.example.com yyy.example.com $hostname for both http and https.

I want to configure the server in such way that all requests to the other domain names will be redirected to xxx.example.com.

Every time I initiate an ssh connection from my Mac to a Linux (Debian) I do get this warning:

No xauth data; using fake authentication data for X11 forwarding.

This also happens for tools that are using ssh, like git or mercurial.

I just want to make a local change to my system in order to prevent this from appearing.

Note: I do have X11 server (XQuartz 2.7.3 (xorg-server 1.12.4)) on my Mac OS X (10.8.1) and it is working properly, I can successfully start clock locally or remotely.

I am currently using ssh to forward few ports like this:

There are 3 machines:

• A-host - machine building the tunnel, one that can access both B and C
• B-host - machine that does not have access to A or C that that needs to be able to connect to C:636
• C-host - machine providing a service on port 636

A-host>ssh -R 636:C-host:636 -R 8080:C-host:8080 root@B-host

The question is how to make this tunnelling be persistent, to restore it even one or another machine is restarted.

I know that this probably require autossh but that't not the entire solution.

Here is the following use case: wanting to ssh server1 in 3 scenarios:

• server1 is on current network, go for it
• server1 is not on current network and you need to use a gateway (gw1) to access it, not a big problem as you can use ~/.ssh/config to define tunnelling for this
• you are on a network that does not allow you to access the first gw1, so you'll use gw2 for that.

Desired outcome: be able to use the same command and be able to connect to the server.

I just want to setup a system wide environment variable, JAVA_HOME for all users, including root user.

Requirements:

• accessible to normal users
• accessible to root
• always loaded, not only for bash (gnome-terminal does not start a bash by default)
• to work on Ubuntu, Debian and optionally Red Hat
• great if addition could be easily scripted

I installed Cygwin and sshd on Windows 7 but I'm not able to make it work.

I followed the guide from http://www.noah.org/ssh/cygwin-sshd.html but I still get:

$ cygrunsrv -S sshd
cygrunsrv: Error starting a service: QueryServiceStatus:  Win32 error 1062:
The service has not been started.

Configuration details:

• Windows 7 in a domain running with a domain user that has local administrative priviledges
• netstat -a -b reports nothing running on port 22
• C:\cygwin\var\log\sshd.log is empty
• After I installed cygwin and sshd, I run ssh-host-config -y
• Already tried to add SYSTEM account to the 3 directories, no change.
• Already tried a full system reboot, no change

Update: Windows logs complains about Possible duplicate cygwin1.dll

How do I activate launchd logging on OS X 10.6?

I added a new daemon that is not starting properly (status is 1).

I want to debug the problem but I was not able to find launchd logs, they are not in /var/log/launchd.log.

On OS X 10.6 you do not have iptables, so what is the full solution for this permanently redirect all request sent to port 80 to another port, like 8080?

It has to be persistent.

I understand that keyservers are using the port 11371 but in many cases you are not allowed to connect to this port and you cannot add

There a many cases when you cannot modify the firewall configuration.

Example command that fails

 gpg --keyserver keyserver.ubuntu.com --recv-keys 0A5174AF

How do you solve this issue?

With yum it was simple: yum list *xxx* but this does not work with apt-get.

How can I do a search like this?

Also, how do I see what package provides a specific command?

How can I install wsgi on CentOS using yum? Is there any repository with mod_wsgi?

I have a 8 CPU VmWare ESX server (3.5) with 4 HDD drives in RAID that is not loaded at all. I enabled SSH and installed mc (midnight commander) in order to be able to copy(clone) virtual machines but I observed that if does copy the files very slow - around 3.5mb/s on local drive.

Why is this happening and how should I solve the issue?