TinyCore RedPill loader (TCRP) - Development release 0.9

[Peter Suh]

20 minutes ago, Head Scratcher said:

I was trying to boot my installation of TCRP (9.2.9) with Friend yesterday and the boot failed after trying to check for Friend updates.  This is when I realized my DNS Server was down.  After bringing my DNS Server back up, I was able to boot Xpenology without any issues.

 

Has anyone else experienced this?  Is the expected behaviour that TCRP with Friend will not boot your server if your DNS Server, the internet or the Friend update location is unavailable?

 

TCRP and TCRP FRIEND always communicate with github during the loader build process.
In the process of downloading Github files, the Internet is absolutely necessary and there should not be any firewall issues. Also, the Internet time and the local time of the PC must not be inconsistent.

 

Separately, TCRP FRIEND always downloads and updates its latest kernel from Github in the process of booting its own loader. Currently, version 0.0.3 is the latest version, and in this process, there are no errors when the above Internet, firewall, and time requirements are met.

[Head Scratcher]

2 hours ago, Peter Suh said:

 

TCRP and TCRP FRIEND always communicate with github during the loader build process.
In the process of downloading Github files, the Internet is absolutely necessary and there should not be any firewall issues. Also, the Internet time and the local time of the PC must not be inconsistent.

 

Separately, TCRP FRIEND always downloads and updates its latest kernel from Github in the process of booting its own loader. Currently, version 0.0.3 is the latest version, and in this process, there are no errors when the above Internet, firewall, and time requirements are met.

I agree it makes sense to need an active internet connection for the initial BUILD of TCRP and FRIEND.
 

However, it doesn’t make sense to me that you should be REQUIRED to have an active internet connection to simply to start or reboot your server.  
 

@pocopico, what do you think?

Edited October 31, 2022 by Head Scratcher

[pocopico]

2 hours ago, Head Scratcher said:

I agree it makes sense to need an active internet connection for the initial BUILD of TCRP and FRIEND.
 

However, it doesn’t make sense to me that you should be REQUIRED to have an active internet connection to simply to start or reboot your server.  
 

@pocopico, what do you think?

 

Internet access is not required for booting but it will be required for updating the TCRP Friend automatically. If internet though is not found it will timeout and continue. If you want autoupdate to be disabled you can do so by modifying user_config.json : 

 

https://github.com/pocopico/tcrpfriend/wiki

 

 

[keasee]

@pocopico sorry Poco,

 

where is the hardware listed you implemented in your driver repositories ?

I'm searching for drivers of an HP NC523SFP  (not intel, it's an Marvell/QLogic cLOM8214) 

 

best keasee

[Vodka2014]

Hi All,

 

I just want to find out if Internet Access to look for updates is disabled by Default in the Loader. I am managed to build DS3622xs and update it to Update 2 no issues at all Manually.

 

Also what is the latest version that is supported, or does it not matter now since the withfriend keeps it uptodate ? 

[Peter Suh]

2 hours ago, Vodka2014 said:

Hi All,

 

I just want to find out if Internet Access to look for updates is disabled by Default in the Loader. I am managed to build DS3622xs and update it to Update 2 no issues at all Manually.

 

Also what is the latest version that is supported, or does it not matter now since the withfriend keeps it uptodate ? 

 

 

 

The reason for this inquiry is that all the issues that you answered me a while ago have been resolved?

 

 

 

 

If it means a separate kernel update of TCRP FRIEND
It is enabled by default.

If you want to disable it, you need to edit user_config.json manually like below.

https://github.com/pocopico/tcrpfriend/wiki
The current latest version is 0.0.3.

https://github.com/pocopico/tcrpfriend

 

The version of the original TCRP is maintained at 0.9.2.9.

https://github.com/pocopico/tinycore-redpill/blob/main/rploader.sh

 

Edited November 1, 2022 by Peter Suh

[Vodka2014]

58 minutes ago, Peter Suh said:

 

 

 

The reason for this inquiry is that all the issues that you answered me a while ago have been resolved?

 

 

 

 

If it means a separate kernel update of TCRP FRIEND
It is enabled by default.

If you want to disable it, you need to edit user_config.json manually like below.

https://github.com/pocopico/tcrpfriend/wiki
The current latest version is 0.0.3.

https://github.com/pocopico/tcrpfriend

 

The version of the original TCRP is maintained at 0.9.2.9.

https://github.com/pocopico/tinycore-redpill/blob/main/rploader.sh

 

 

Thanks, but this does not really answer my Question i am asking. 

 

So all Future Updates will just work correct ? 

 

Also why is the DSM Checks Disabled ? 

[Peter Suh]

56 minutes ago, Vodka2014 said:

 

Thanks, but this does not really answer my Question i am asking. 

 

So all Future Updates will just work correct ? 

 

Also why is the DSM Checks Disabled ? 

 

If you keep the TCRP FRIEND loader, it will be ready for any future updates.

 

Disabling the update check is your choice, but it doesn't seem to have anything to do with the availability of future updates.

[Vodka2014]

2 minutes ago, Peter Suh said:

 

If you keep the TCRP FRIEND loader, it will be ready for any future updates.

 

Disabling the update check is your choice, but it doesn't seem to have anything to do with the availability of future updates.

 

Thanks alot this is what i was looking for so nothing the user needs to do it will just Auto Update

 

for the DSM Checks This is what i mean. https://imgur.com/a/oqhqEa2

[Peter Suh]

5 minutes ago, Vodka2014 said:

 

Thanks alot this is what i was looking for so nothing the user needs to do it will just Auto Update

 

for the DSM Checks This is what i mean. https://imgur.com/a/oqhqEa2

 

In the past, loaders (except jun's mod) had a problem of falling into a recovery mode loop when auto-updating DSM.
So most of the loader builds would have recommended this third option.
And, this third option is not loader-guided
You may have asked a question during the process of creating a user account after installing DSM.
This is what you chose. Think carefully.

[Vodka2014]

Just now, Peter Suh said:

 

In the past, loaders (except jun's mod) had a problem of falling into a recovery mode loop when auto-updating DSM.
So most of the loader builds would have recommended this third option.
And, this third option is not loader-guided
You may have asked a question during the process of creating a user account after installing DSM.
This is what you chose. Think carefully.

 

All good will double check this all again. But thanks for the help and support really appreciate it. 

[pocopico]

On 10/31/2022 at 5:23 PM, keasee said:

@pocopico sorry Poco,

 

where is the hardware listed you implemented in your driver repositories ?

I'm searching for drivers of an HP NC523SFP  (not intel, it's an Marvell/QLogic cLOM8214) 

 

best keasee

Hi, can you please share the PCIID of your device ? an lspci -nnq from TCRP would do fine.

 

EDIT : it looks though like a qlcnic adapter :

 

https://github.com/pocopico/rp-ext/raw/main/qlcnic/rpext-index.json

 

 

Edited November 1, 2022 by pocopico

[keasee]

On 11/1/2022 at 6:49 PM, pocopico said:

Hi, can you please share the PCIID of your device ? an lspci -nnq from TCRP would do fine.

 

EDIT : it looks though like a qlcnic adapter :

 

https://github.com/pocopico/rp-ext/raw/main/qlcnic/rpext-index.json

 

 

no, not now -  because it should be a purchase decision, I want to buy it ... 

 

HPE NC523SFP 10Gb 2-port SFP+ Server Adapter

maybe: QLE8242 Dual Port 10GbE Network adapter ?

 

General Specifications : 
Onboard memory 256MB
Data rate10,000 Mbps per port, (40 Gbps aggregate full duplex, dual ports)
Data Path8 Lane (x8) PCI Express, compatible with x4 and x8 bus widths
Dimensions(LxW)6.6 x 2.7 in (16.7 x 6.8 cm) without bracketPower and Environmental Specifications Operating Temperature32° to 131° F (0° to 55° C)Humidity10% to 90% non-condensing Non-operatingTemperature-40° to 158° F (-40° to 70° C)Humidity5% to 95%

NC523SFP is an eight lane (x8) PCI Express (PCIe) 10 Gigabit network solution offering superior bandwidth in a ProLiant Ethernet adapter. This dual port PCI Express Gen 2 adapter supports SFP+ (Small Form-factor Pluggable) connectors, requiring either Direct Attach Cable (DAC) for copper environments, or fiber transceivers and fiber cables for fiber optic environments

 

 

Edited November 2, 2022 by keasee

[pocopico]

15 minutes ago, keasee said:

no, not now -  because it should be a purchase decision, I want to buy it ... 

 

HPE NC523SFP 10Gb 2-port SFP+ Server Adapter

maybe: QLE8242 Dual Port 10GbE Network adapter ?

 

General Specifications : 
Onboard memory 256MB
Data rate10,000 Mbps per port, (40 Gbps aggregate full duplex, dual ports)
Data Path8 Lane (x8) PCI Express, compatible with x4 and x8 bus widths
Dimensions(LxW)6.6 x 2.7 in (16.7 x 6.8 cm) without bracketPower and Environmental Specifications Operating Temperature32° to 131° F (0° to 55° C)Humidity10% to 90% non-condensing Non-operatingTemperature-40° to 158° F (-40° to 70° C)Humidity5% to 95%

NC523SFP is an eight lane (x8) PCI Express (PCIe) 10 Gigabit network solution offering superior bandwidth in a ProLiant Ethernet adapter. This dual port PCI Express Gen 2 adapter supports SFP+ (Small Form-factor Pluggable) connectors, requiring either Direct Attach Cable (DAC) for copper environments, or fiber transceivers and fiber cables for fiber optic environments

 

 

 

Hi, @arTii has found a nice and inexpensive solution here :

 

 

 

[pocopico]

19 minutes ago, keasee said:

no, not now -  because it should be a purchase decision, I want to buy it ... 

 

HPE NC523SFP 10Gb 2-port SFP+ Server Adapter

maybe: QLE8242 Dual Port 10GbE Network adapter ?

 

General Specifications : 
Onboard memory 256MB
Data rate10,000 Mbps per port, (40 Gbps aggregate full duplex, dual ports)
Data Path8 Lane (x8) PCI Express, compatible with x4 and x8 bus widths
Dimensions(LxW)6.6 x 2.7 in (16.7 x 6.8 cm) without bracketPower and Environmental Specifications Operating Temperature32° to 131° F (0° to 55° C)Humidity10% to 90% non-condensing Non-operatingTemperature-40° to 158° F (-40° to 70° C)Humidity5% to 95%

NC523SFP is an eight lane (x8) PCI Express (PCIe) 10 Gigabit network solution offering superior bandwidth in a ProLiant Ethernet adapter. This dual port PCI Express Gen 2 adapter supports SFP+ (Small Form-factor Pluggable) connectors, requiring either Direct Attach Cable (DAC) for copper environments, or fiber transceivers and fiber cables for fiber optic environments

 

 

 

Personaly i cannot comment on 10Gb card as i havent install any on physical  DSM yet. What i recommend though is to buy a fanless one as the fans on the ones that i have, are very noisy on my home lab.

[AngryPapa]

Hello everyone, I've run into a problem. E10G21-F2 network card does not work. Not visible in dsm 7.1.0

[Peter Suh]

7 hours ago, AngryPapa said:

Hello everyone, I've run into a problem. E10G21-F2 network card does not work. Not visible in dsm 7.1.0

 

I have already posted an answer to this post looking for a solution to the issue. please check.

 

 

[minigranis]

I am running TCRP 0.9 on ESXI and everything is fine, until I map a USB drive to my TCRP virtual machine in ESXI, then it wont boot correctly.

USB boot is disabled in BIOS (EFI) and my VM boots into GRUB just fine but no matter what GRUB option I choose it wont boot, see supplied screenshot.
If I remove the USB within ESXI all works as expected, adding the USB drive when Synology has booted also works fine.
One note, I dont know if it matters, but I had no USB drives attached to the VM when I "created" the TCRP loader.

As of now if I get a restart of my VM I cant boot into TCRP/Synology without first removing the USB drives in ESXI. 

 

Any hints if this is a user error or something that could be fixed ?

 

[Peter Suh]

2 hours ago, minigranis said:

I am running TCRP 0.9 on ESXI and everything is fine, until I map a USB drive to my TCRP virtual machine in ESXI, then it wont boot correctly.

USB boot is disabled in BIOS (EFI) and my VM boots into GRUB just fine but no matter what GRUB option I choose it wont boot, see supplied screenshot.
If I remove the USB within ESXI all works as expected, adding the USB drive when Synology has booted also works fine.
One note, I dont know if it matters, but I had no USB drives attached to the VM when I "created" the TCRP loader.

As of now if I get a restart of my VM I cant boot into TCRP/Synology without first removing the USB drives in ESXI. 

 

Any hints if this is a user error or something that could be fixed ?

 

 

Have you ever used M SHELL?
As far as I know pocopico's TCRP is not using msdos3.
Recently 3 days ago, by the advice of pocopico
I adjusted the boot partition of M SHELL to the 3rd one.
https://github.com/PeterSuh-Q3/redpill-load/commit/58ac1575cd3558f9a554b21978f07bd5c187dc5a

What model was built and when?

[pocopico]

10 hours ago, minigranis said:

I am running TCRP 0.9 on ESXI and everything is fine, until I map a USB drive to my TCRP virtual machine in ESXI, then it wont boot correctly.

USB boot is disabled in BIOS (EFI) and my VM boots into GRUB just fine but no matter what GRUB option I choose it wont boot, see supplied screenshot.
If I remove the USB within ESXI all works as expected, adding the USB drive when Synology has booted also works fine.
One note, I dont know if it matters, but I had no USB drives attached to the VM when I "created" the TCRP loader.

As of now if I get a restart of my VM I cant boot into TCRP/Synology without first removing the USB drives in ESXI. 

 

Any hints if this is a user error or something that could be fixed ?

 

you can press ‘e’ at GRUB and test hd1 , hd2 etc until you find the right drive. Then for a permanent solution you can edit the grub.cig 

[minigranis]

16 hours ago, Peter Suh said:

 

Have you ever used M SHELL?
As far as I know pocopico's TCRP is not using msdos3.
Recently 3 days ago, by the advice of pocopico
I adjusted the boot partition of M SHELL to the 3rd one.
https://github.com/PeterSuh-Q3/redpill-load/commit/58ac1575cd3558f9a554b21978f07bd5c187dc5a

What model was built and when?

 

Thanks for jumping in, yes I have used M SHELL  

This loader is an upgrade from the released TCRP loader and I used M SHell to build that.

Model is DS3622xs+ and build with TCRP (about version 0.6) M shell.

 

[minigranis]

2 hours ago, minigranis said:

Thanks for jumping in, yes I have used M SHELL  

This loader is an upgrade from the released TCRP loader and I used M SHell to build that.

Model is DS3622xs+ and build with TCRP (about version 0.6) M shell.

 

Follow up - tried to generate a brand new TCRP image and same issue occurs as soon as I add a USB drive.
Editing the GRUB as @pocopico suggested works but it would be neat to find a more flexible solution if possible, now I have to remember to edit the GRUB if I add / remove USB drives. Can that be achieved ?

 

[Peter Suh]

6 hours ago, minigranis said:

 

Follow up - tried to generate a brand new TCRP image and same issue occurs as soon as I add a USB drive.
Editing the GRUB as @pocopico suggested works but it would be neat to find a more flexible solution if possible, now I have to remember to edit the GRUB if I add / remove USB drives. Can that be achieved ?

 

 

 

There is one more thing I want to check.

Is it baremetal or VM?
Legacy BOOT or UEFI in BIOS?
Normally, it is correct to load kernel files from the 3rd partition of TCRP FRIEND from the MSDOS3 partition.

I'd like to know more about what's special about your motherboard and USB stick.
Is it correct to use only one USB stick?

If you want to permanently fix the MSDOS# value you find
Insert a USB stick into MS WINDOWS and change the MSDOS# value by editing the grub.cfg file located in the first partition.

[minigranis]

5 hours ago, Peter Suh said:

Is it baremetal or VM?
Legacy BOOT or UEFI in BIOS?
Normally, it is correct to load kernel files from the 3rd partition of TCRP FRIEND from the MSDOS3 partition.

I'd like to know more about what's special about your motherboard and USB stick.
Is it correct to use only one USB stick?

 

 

Its a VM, I am running a HP Gen8 Microserver with VMWare/ESXI, so I boot from a VMDK. No other USB drives  apart from those I "want to use in Synology", in total I have 3 USB drives (real HDDs) that I use in Synology.

 

EFI boot, I have also tried with Legacy with the same problem (IIRC)

I have edited the grub.cfg so it will boot after restart/powerfailure etc for now - but as I said, editing the grub.cfg is a "bit to static" IMHO.

 

Thanks for helping out !

[pocopico]

14 hours ago, minigranis said:

 

Follow up - tried to generate a brand new TCRP image and same issue occurs as soon as I add a USB drive.
Editing the GRUB as @pocopico suggested works but it would be neat to find a more flexible solution if possible, now I have to remember to edit the GRUB if I add / remove USB drives. Can that be achieved ?

 

 

Yes, actualy we can instruct grub to search for the UUID of the disk. If you replace the set root line with :

 

search --set=root --fs-uuid BF4E-ADE7 --hint hd0,msdos3 (or 1 depends on your line)

 

Doing so the grub root will be always set to the UUID of the partition UUID.

 

/dev/sda2: SEC_TYPE="msdos" UUID="BF6B-61D1" 
/dev/sda3: UUID="6234-C863" 
/dev/sda1: SEC_TYPE="msdos" UUID="BF4E-ADE7" 

 

EDIT: Modified rploader.sh to now set root to the UUID instead. I had this plan in the past that somehow got stuck in between other issues. You can try with latest release v0.9.3.0

 

 

Edited November 15, 2022 by pocopico