Installed TC on VMWARE 6.7 with open-vm-tools but tools are not running

[LeonStraathof]

Installed TC on VMWare ESX 6.7 and installed open-vm-tools.tcz. It is also in the onboot.lst but when rebooting host still says vmtools are not running.  I am running TC without GUI so shell only . I want to make sure that TC loaded the optimized vmxnet3 driver since i want to use the TC install as a very little bridge.

How to check on TC if vmtools are running as i suspect ESX just not detecting right. I am not sure because on other linux distro's ESX always detects correct that tools is running.

Also some help with installing a bridge extension would be nice because alle extensions i found online are not installing with tce-load which puzzles me because i thought there was one uniform repository for TC??

I am running TC version 11.1

[andyj]

The vmtoolsd service does not start automatically. Read the info file that comes with the open-vm-tools extension for more information. You will need to add the services you wish to start to /opt/bootlocal.sh. Try one of the following examples:

Code: [Select]

# start all installed services
for a in $(find -L /usr/local/etc/init.d -type f); do
      [ -x $a ] && $a start &
done
or

Code: [Select]

# start only listed services
for a in gpm openssh open-vm-tools bind php-fpm nginx httpd; do
        [ -x /usr/local/etc/init.d/$a ] && /usr/local/etc/init.d/$a start &
done


[LeonStraathof]

Ok thanks i should have been reading more to understand the working of TC a bit more. Did take it for regular linux but now i understand some things are not configured out of the box. Like services.

I did the following:
sudo vi /opt/bootlocal.sh

added the following hardcoded line instead of the fancy example:
/usr/local/etc/init.d/open-vm-tools start

esc :x to save and exit.
checked content of the file /opt/bootlocal.sh with cat and all was ok.
checked content of the file /opt/.filetool.lst which has the default content:
opt
home

made file persistent (yes did some reading).
sudo filetool.sh -b
it says it is backing up the file but the done message is never appearing??

sudo reboot

And now all my changes in the /opt/bootlocal.sh are gone.

I think i followed the guides ok why is the filetool not making a correct backup??

[LeonStraathof]

I looked at the output of the filetool.sh -b statement. It created a new /mnt/sda1/tce/mydata.tgz file (timestamp changes each time i backup) but filesize is always 0 bytes??

[andyj]

With filetool.sh no news is bad news. If it doesn't say "Done." it failed. Something you are trying to back up is unreadable or unreachable. Try filetool.sh with the -v verbose option and see if that gives you a hint to what it doesn't like.

[LeonStraathof]

Thanks again for the help, the verbose option gave me a silly answer that made a lot of sense (disk full). Most OS variants give other messages/problems when your disk is full, but i guess the stateless nature of TC prevents those other problems. I still like TC although the learning curve is steep and information is scattered. The fact that the wiki seems to be down forever does not help to get to information (User authentication is temporarily unavailable. If this situation persists, please inform your Wiki Admin). But the forum seems to bee a great place with a lot of dedicated people. I just wish there was more online reading material for people who are starting with TC.

I have 2 other questions:
1 What are the official minimal specs for TC because i found out the hard way that it is not the advertised memory footprint that is on the site. Sure when it is running the consumption is low, but to boot it (get it to a running state) takes more so sounds bit of false advertising to me. I can do trial and error to figure out the approx. minimal requirements but if anyone can tell would be nice.

2 I still want to configure this as a bridge and could not figure out if there is a ready made TCE. To explain in more detail what i am trying to build:
I want a small as possible ESX VM that boots stateless (not interested in anything being saved of the user session) and acts as a bridge between 2 network adapters. One network adapter is a virtual vmxnet3 adapter the other one will be a physical adapter that is connected to the VM via passthrough. The physical adapter will in most cases be one of the many Realtek gigabit chipset types. Since VMware ESX 7 does not support linux drivers anymore only native ESX drivers, using a dedicated VM as a bridge is my solution for using not supported nic's on ESX. I got it working with a Debian install that i set as autoboot first VM when the ESX host starts and it works like a charm. The reason i looked at TC to replace the Debian solution is size (the smaller the better) and the fact that TC is stateless (i want it to boot clean each time it boots and no live kernel updates and other stuff changing the machine over time). Any idea how i could build something that would just bridge automatically when it sees 2 network adapters?

[Rich]

Hi LeonStraathof

... I still like TC although the learning curve is steep and information is scattered. The fact that the wiki seems to be down forever does not help to get to information (User authentication is temporarily unavailable. If this situation persists, please inform your Wiki Admin).  ...

There is a mirror of the Wiki here:
http://wiki.tinycorelinux.net/doku.php?id=wiki:start
You could take an hour and read this very fine book here:
http://tinycorelinux.net/corebook.pdf
And you probably already found the FAQ:
http://tinycorelinux.net/faq.html

... 1 What are the official minimal specs for TC because i found out the hard way that it is not the advertised memory footprint that is on the site. ...

That number changes over time as new releases come out. Resources are not dedicated to measuring it since the
issue almost never comes up anymore.

... I can do trial and error to figure out the approx. minimal requirements ...

Since you want to do more than simply boot, you will have to do that anyway. Any "minimum RAM numbers" assume
a bare system with no applications installed and running.

... 2 I still want to configure this as a bridge and could not figure out if there is a ready made TCE. ...
 ... I got it working with a Debian install that i set as autoboot first VM when the ESX host starts and it works like a charm. ...

What application(s)/daemons did you start under Debian to accomplish your goal? That information can be used to see
if the TCEs you need exist. Don't spout out systemd (sysconfig) commands. They don't exist in Tinycore.

[LeonStraathof]

Hey Rich, thanks for the reply, and pointing me to the mirror of the wiki that really helps me a lot. Also thanks for pointing out the book in pdf format, already made corrections to my machine thanks to info in the book (made a vm with the HD connected to ide as claimed to be the best solution for ESX in some forum posts, but the book stated that is should use paravirtualized scsi as thats is the supported solution) This small change made the vm boot even faster so i am a very happy camper.

About the minimal specs, it aint that big of a issue because this must by far be the smallest distro anyway. It's just that the numbers on the website seem to be from a ancient version. It's like Microsoft stating that they have a Windows version that can run on 256kb of memory, which is no lie in 1985 they had Windows 1.0 that did exactly that.

Already started digging in making your own extentsions so even if the extension i need does not exist yet i guess i could make it myself. I will retrace my steps on the Debian machine first to give a complete answer here first. Because if it already exists no need to reinvent the wheel.

Still the way this OS behaves is fundamently different then other linux variants. The concept that it creates an running environment from essentially merging the content of several archive files is very smart. However when i look at what most people are trying to do with this OS is making a machine that is not completly stateless which is the opposite of what this OS is very good at.

[Rich]

Hi LeonStraathof
This may be of use to you:
http://tinycorelinux.net/11.x/x86/tcz/iproute2.tcz.info
There's a 64 bit version available too. You didn't specify which architecture you were interested in.

There's a brief answer here:
https://unix.stackexchange.com/questions/255484/how-can-i-bridge-two-interfaces-with-ip-iproute2

This fellow, discontent with the iproute2 documentation, set up his own documentation page with examples:
https://baturin.org/docs/iproute2/

... Still the way this OS behaves is fundamently different then other linux variants. The concept that it creates an running environment from essentially merging the content of several archive files is very smart. However when i look at what most people are trying to do with this OS is making a machine that is not completly stateless which is the opposite of what this OS is very good at.

So, your one of those guys who pays attention, reads documentation, and takes the time to comprehend what's going on?
I see we're going to have to keep an eye on you.

[andyj]

I have one of what you are looking for, but my gateway VM is also diskless, I have another VM running dhcp/bootp and lighttpd to boot it using PXE. It's as stateless as can be, I don't know how I could save anything for persistence even if I wanted to. You say bridging which is not the same as routing and firewalling. Are the two NICs on the same network/subnet or different networks/subnets?

[LeonStraathof]

Yeah stateless is great when you have a task for a machine it just has to do until the end of times. Build once and after that just use. Since VMware dropped support of Linux network drivers which everybody dislikes but for VMware's point of view is understandable. I got the idea to just let the network interfaces live in it's own habitat with a real Linux kernel. So for the the question if routing is the same as bridging the answer is no it is not the same. When you bridge 2 interfaces both ends connect 2 networks 2 become 1 subnet. So if on one side of your bridge you have a dhcp server it can serve ip addresses to machines at the other side of the bridge. Most used home user scenario is when you put your isp's modem in bridge and putting your own router behind it. The wan port of your own router will get public internet ip address. The static configured ip of the bridge machine itself can still be used to connect to the bridge machine for configuration if needed. I don't think i will need configuration trough network though as i plan to build everything preconfigured in a stateless machine.

On Debian i build something by doing these easy steps:

Code: [Select]

apt install bridge-utils

Edit: /etc/network/interfaces
 auto lo br0
 iface lo inet loopback

 iface eth0 inet manual
 iface eth1 inet manual

 iface br0 inet static
    bridge_ports eth0 eth1
        address 192.168.111.1
        netmask 255.255.255.0
        gateway 192.168.111.3
In this example the address for connecting to the Debian machine trough ssh to configure something is 192.168.111.1 it is not the address you direct other traffic to. All the traffic in the subnet 192.168.111.1/24 is transparent as far as the bridge is concerned. You also see that although the bridge has 2 interfaces they only have 1 ip adress for the whole bridge.

Still reading documentation though trying to find how to configure network interfaces in TC.

On Debian you can verify the bridge status with brctl show
did a tce-load net-bridging-5.4.3-tinycore  but that does not seem to have brctl in it but it looks like the core items for bridging are in there.

Yeah i am on 32bit, have no preference yet for 32 vs 64 bit. Probably build both if i get things running and choose which i am going to use after some performance testing.

    [EDIT]: Added code tags.  Rich

[Rich]

Hi LeonStraathof
Please use  Code Tags  when posting commands and responses seen in a terminal. To use  Code Tags  click on the  #  icon
above the reply box and paste your text between the  Code Tags  as shown in this example:

[code][   36.176529] pcm512x 1-004d: Failed to get supply 'AVDD': -517
[   36.176536] pcm512x 1-004d: Failed to get supplies: -517
[   36.191753] pcm512x 1-004d: Failed to get supply 'AVDD': -517[/code]

It will appear like this in your post:

Code: [Select]

[   36.176529] pcm512x 1-004d: Failed to get supply 'AVDD': -517
[   36.176536] pcm512x 1-004d: Failed to get supplies: -517
[   36.191753] pcm512x 1-004d: Failed to get supply 'AVDD': -517
Code Tags  serve as visual markers between what you are trying to say and the information you are posting. They also preserve
spacing so column aligned data displays properly.

[Rich]

Hi LeonStraathof

... Still reading documentation though trying to find how to configure network interfaces in TC. ...

This is one way of doing it:

Code: [Select]

tc@E310:~$ cat /opt/eth0.sh
#!/bin/sh
pkill udhcpc
ifconfig eth0 192.168.1.35 netmask 255.255.255.0 broadcast 192.168.1.255 up
route add default gw 192.168.1.1
echo nameserver 192.168.1.1 > /etc/resolv.conf
echo nameserver 68.237.161.12 >> /etc/resolv.conf
tc@E310:~$/opt/eth0.sh  gets called from  /opt/bootlocal.sh.

... did a tce-load net-bridging-5.4.3-tinycore  but that does not seem to have brctl in it but it looks like the core items for bridging are in there. ...

Those are kernel modules. I don't see any extensions in the repository that depend on them yet.

[Juanito]

There’s bridge-utils in the x86_64 repo.

[LeonStraathof]

Thanks Rich for pointing out how to configure network settings.

Juanito, thanks for mentioning that it is in the 64 bit build. guess i will be switching to that instead of the 32bit build. Any reason why there is so many differentiation between all repository folders of all builds? Or is it just because of development capacity and what was asked to be build first?