Experiences with test version of Debian Edu v12
Hi folks, I have repeated some of the tests I did with Debian Edu v11 the other day. It has been suggested, that I should repeat those tests with the current test version of Debian Edu. These are my findings: I have set up a new server using debian-edu-12.5.0-amd64-BD-1.iso. I was a little surprised how fundamental the changes are compared to my old Debian Edu based on Buster. The old system had its up and downs but it was actually working rather well for us. According to the manual the client is built automatically. I can confirm this and in my case client data is built in /srv/ltsp/x2go-bare-amd64. The PXE menu remains rather empty though, e.g. not a single LTSP entry is added. Btw, the PXE menu now really looks ugly compared to what we had in earlier versions of Debian Edu. Anyway, this is just cosmetics. Consequently I decided to build the clients afresh using the following command: debian-edu-ltsp-install --diskless_workstation yes --thin_type bare The update failed during the process of creating the thin client and the errors were nicely listed in /srv/ltsp/x2go-bare-amd64/debootstrap/debootstrap.log. The errors listed didn't make much sense to me and smelled like bugs to me. I decided consequently to delete /srv/ltsp/x2go-bare-amd64/ and start the whole build afresh instead of updating it. This worked just fine and the PXE menu showed the following 3 entries: 1. Plain X2Go Thin Client (64-Bit) 2. Diskless Workstation (64-Bit, server's SqashFS image) 3. Plain X2Go Thin Client (64-Bit, NFS rootfs) So far I could not find out the difference between entry 1 and 3. They may actually point to the same image. I didn't check that yet. I have various clients at my disposal. "diskless workstation" works just fine for newer hardware but my older Toughbooks no longer boot and I'm getting the following message: This Kernel requires an x86-64 CPU, but only detected an i686 CPU. Unable to boot - please use a kernel appropriate for your CPU. AFAIK this could rather easily be fixed in earlier versions of Debian Edu, e.g. by simply creating a i386-chroot, which then worked just fine. Alternatively the server could be build using the i386 version. "thin client" works just fine on newer hardware. On my Toughbooks it also fails with the same message as above. "thin client" is still rather mysterious to me. I'm now getting a somehow strange looking screen dividing login screen instead of the old full screen login which basically looked just the same as the "diskless" one. Anyway, I didn't find time to do more tests. There are issues however updating screen changes. Some parts fail to update and only do so when hover with the mouse over that area. To make my computer lab run with older Toughbooks is important to us. That's all I have for now. Thanks in advance for some input on the issues. Kind regards, Roman
Experiences with Debian Edu v11
Hi folks, I would like to share some experiences I made last week with Debian Edu v11. I have set up a new server using firmware-edu-11.8.0-amd64-BD-1.iso. I decided to go for the firmware version to avoid problems with missing drivers I often had in the past delaying me too much. I was a little surprised how fundamental the changes are compared to buster. The old system had its up and downs but it was actually working rather well for us. According to the manual the client is built automatically. I can confirm this and in my case client data is built in /srv/ltsp/x2go-bare-amd64. The PXE menu remains rather empty though, e.g. not a single LTSP entry is added. Btw, the PXE menu now really looks ugly compared to what we had in earlier versions of Debian Edu. Anyway, this is just cosmetics. Consequently I decided to build the clients afresh using the following command: debian-edu-ltsp-install --diskless_workstation yes --thin_type bare The update failed during the process of creating the thin client and were nicely listed in /srv/ltsp/x2go-bare-amd64/debootstrap/debootstrap.log. The errors listed didn't make much sense to me and smelled like bugs to me. I decided consequently to delete /srv/ltsp/x2go-bare-amd64/ and start the whole build afresh instead of updating it This worked just fine and the PXE menu showed the two following entries: • Diskless Workstation (64-Bit, server's SquashFS image) • Plain X2Go Thin Client (64-Bit, NFS rootfs) I have various clients at my disposal. "diskless workstation" works just fine for newer hardware but my older Toughbooks no longer boot and I'm getting the following message: This Kernel requires an x86-64 CPU, but only detected an i686 CPU. Unable to boot - please use a kernel appropriate for your CPU. AFAIK this could rather easily be fixed in earlier versions of Debian Edu simply by creating a i386-chroot, which then worked just fine. "thin client" is still rather mysterious to me. I'm now getting a somehow strange looking screen dividing login screen instead of the old full screen login which basically looked just the same as the "diskless workstation" one. Anyway, I didn't find time to do more tests. To make my computer lab run with older Toughbooks is important to us. I consequently decided to switch to another iso image being firmware-edu-11.9.0-i386-BD-1.iso. I was hoping that using the i386 version instead of amd64 would allow my Toughbooks to boot just fine. This time I got the following two PXE menu entries: • Diskless Workstation (32-Bit, server's SquashFS image) • Plain X2Go Thin Client (32-Bit, NFS rootfs) The experiences are not really better than with amd64. I'm getting weird messages when booting "diskless workstation" on newer hardware, e.g. [FAILED] Failed to start BIND Domain Name Server [FAILED] Failed to start Samba SMB Daemon [FAILED] Failed to start Kerberos 5 Key Distribution Center [FAILED] Failed to start Squid Web Proxy Server. It appears that I'm having an issue with a buggy script creating my SquashFS image. Anyway, I'm getting to the login screen and I can login but the login screen reappears shorty after again. The Toughbooks still don't boot using "diskless workstation". This time I got the following message: WARNING: PAE disabled. Use parameter 'forcepae' to enable at your own risk! This kernel requires the following features not present on the CPU: pae Unable to boot - please use a kernel appropriate for your CPU. I tried to update my client builds but I failed to do so. The server is unable to access the internet. IP's can be pinged but domain names cannot. That's all I have for now. Thanks in advance for some input on the many issues. ;) Kind regards, Roman
Re: User login issue
Hi folks,
Yesterday, I came across the following entry in /var/log/auth.log:
Feb 6 11:03:38 tjener su: pam_krb5(su:auth): (user roman) credential
verification failed: Cannot find key for host/tjener.intern@INTERN kvno 16 in
keytab
I also had a closer look at the following script:
/usr/share/debian-edu-config/tools/copy-host-keytab
This then lead me to the solution of my authentication problem.
My file /etc/krb5.keytab was missing many entries preventing successful user
logins. Executing the script fixed this finally.
Kind regards,
Roman
> On 01/07/2024 11:07 AM GMT roman.me...@gismap.ch wrote:
>
>
> Hi folks,
>
> Maybe the following is helping to narrow things down?
>
> I checked on /var/log/auth.log today and I'm getting the following upon
> trying to login as user mm in the console:
>
> Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
> 26}) 10.0.2.2: NEEDED_PREAUTH: mm@INTERN for krbtgt/INTERN@INTERN, Additional
> pre-authentication required
> Jan 7 11:04:34 tjener krb5kdc[2232]: preauth (encrypted_timestamp) verify
> failure: Preauthentication failed
> Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
> 26}) 10.0.2.2: PREAUTH_FAILED: mm@INTERN for krbtgt/INTERN@INTERN,
> Preauthentication failed
> Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
> 26}) 10.0.2.2: NEEDED_PREAUTH: mm@INTERN for krbtgt/INTERN@INTERN, Additional
> pre-authentication required
> Jan 7 11:04:34 tjener krb5kdc[2232]: preauth (encrypted_timestamp) verify
> failure: Preauthentication failed
> Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
> 26}) 10.0.2.2: PREAUTH_FAILED: mm@INTERN for krbtgt/INTERN@INTERN,
> Preauthentication failed
> Jan 7 11:04:34 tjener login[17928]: pam_krb5(login:auth): authentication
> failure; logname=mm uid=0 euid=0 tty=/dev/tty1 ruser= rhost=
> Jan 7 11:04:34 tjener login[17928]: pam_unix(login:auth): authentication
> failure; logname=LOGIN uid=0 euid=0 tty=/dev/tty1 ruser= rhost= user=mm
> Jan 7 11:04:38 tjener login[17928]: FAILED LOGIN (1) on '/dev/tty1' FOR
> 'mm', Authentication failure
>
> Kind regards,
> Roman
Re: User login issue
Hi folks,
Maybe the following is helping to narrow things down?
I checked on /var/log/auth.log today and I'm getting the following upon trying
to login as user mm in the console:
Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
26}) 10.0.2.2: NEEDED_PREAUTH: mm@INTERN for krbtgt/INTERN@INTERN, Additional
pre-authentication required
Jan 7 11:04:34 tjener krb5kdc[2232]: preauth (encrypted_timestamp) verify
failure: Preauthentication failed
Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
26}) 10.0.2.2: PREAUTH_FAILED: mm@INTERN for krbtgt/INTERN@INTERN,
Preauthentication failed
Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
26}) 10.0.2.2: NEEDED_PREAUTH: mm@INTERN for krbtgt/INTERN@INTERN, Additional
pre-authentication required
Jan 7 11:04:34 tjener krb5kdc[2232]: preauth (encrypted_timestamp) verify
failure: Preauthentication failed
Jan 7 11:04:34 tjener krb5kdc[2232]: AS_REQ (8 etypes {18 17 20 19 16 23 25
26}) 10.0.2.2: PREAUTH_FAILED: mm@INTERN for krbtgt/INTERN@INTERN,
Preauthentication failed
Jan 7 11:04:34 tjener login[17928]: pam_krb5(login:auth): authentication
failure; logname=mm uid=0 euid=0 tty=/dev/tty1 ruser= rhost=
Jan 7 11:04:34 tjener login[17928]: pam_unix(login:auth): authentication
failure; logname=LOGIN uid=0 euid=0 tty=/dev/tty1 ruser= rhost= user=mm
Jan 7 11:04:38 tjener login[17928]: FAILED LOGIN (1) on '/dev/tty1' FOR 'mm',
Authentication failure
Kind regards,
Roman
Re: User login issue
Hi Mike, > This very likely means that your Kerberos layer / service stack is broken. > > Do you have libpam-krb5 installed on TJENER? (That would be an easy solution). Nope, it was not installed. Maybe my legacy installation is not needing it? I installed it but things did not improve. > Does the new user object in LDAP have krb* LDAP attributes? Yep, I found 9 entires: krbPrincipalName: mm@INTERN krbPwdPolicyReference: cn=users,cn=INTERN,cn=kerberos,dc=skole,dc=skolelinux,dc=no krbLoginFailedCount: 0 krbTicketFlags: 128 krbPrincipalKey:: AwIBAqMDAgEBpIICPjCCAjowVKAHMAWgAwIBAKFJMEeg[...] krbPasswordExpiration: 1970010100Z krbLastPwdChange: 20240105153122Z krbExtraData:: AALKIJhlcm9vdC9hZG1pbkBJTlRFUk4A krbExtraData:: AAgBAA== > If you launch kadmin.local and then enter "list_principals": do any > Kerberos principals (users and/or hosts and/or services) get shown? Do > the user accounts that fail login get listed by this? Yep, they get all nicely listed. > If the new LDAP users don't get listed, try "add_princ -policy users > " and try login from another tty. > > If the new LDAP users get listed, try to set their password using "cpw ". I did this but the user still can't login. > Please also let me/us know what versions of Debian Edu you have > installed (11 or 12)? This one is my personal debian edu workstation and testserver. It's rather legacy and still on 10 (buster) with GOsa 2.7.4. > If 12, have you upgraded to latest package > versions? There was a bug in Debian Edu 12's debian-edu-config that > only got resolved recently: > > ``` > debian-edu-config (2.12.41~deb12u1) bookworm; urgency=medium > >* Upload to bookworm. > > -- Mike Gabriel Sun, 03 Dec 2023 08:45:42 +0100 > > debian-edu-config (2.12.41) unstable; urgency=medium > >[ Guido Berhoerster ] >* gosa-sync: Decode the user password which GOsa substitutes base64 > encoded. > This fixes a bug where the user password could not be set or changed. > (related to #1052159). > > -- Mike Gabriel Fri, 01 Dec 2023 21:44:38 +0100 > ``` > > This fix in d-e-c goes together with a fix in gosa: d-e-c? > ``` > gosa (2.8~git20230203.10abe45+dfsg-1+deb12u2) bookworm; urgency=medium > >[ Daniel Teichmann ] >* debian/patches: > [...] > + Add 1044_fix-class-ldap-serialization.patch which fixes a few bugs >regarding serialization. This especially fixes setting LDAP > userPassword >attribute types via GOsa². (Closes: #1052159). > + Add 1045_fix-posixaccount-shadowExpire.patch which fixes shadowExpire >always being set to 0. (User can't login then). (Closes: #1053806). > >[ Guido Berhoerster ] >* debian/patches: > [...] > >[ Mike Gabriel ] >* debian/patches: > [...] > > -- Mike Gabriel Sun, 03 Dec 2023 08:16:31 +0100 > > If you Debian Edu 12, simply upgrading d-e-c and gosa to the > referenced versions should help. > > Mike Kind regards, Roman
Re: User login issue
Hi Mike, > One thing caught me at first glance that is strange: why does > ldapsearch try GSS auth although you requested simple_bind > authentication? Hmmm... is that -x in your quoted command really > starting with a '-' dash / minus sign? It looks longer (like an > )... Yep, you were right! Thanks! > If you create a new user account via GOSa can you login with that > account then? I can create a new user but the behavior is the same: I cannot login on the server. Login into GOsa2 works fine. > Greets, > Mike It also came to my mind that GOsa2 requested for a MAC address for tjener when I tried to add a DHCP group, which I did. It felt somehow wrong though since tjener normally has two interfaces. Anyway, I also deleted the entry using commandline which did not improve things. I'm not sure this is related though. Kind regards, Roman
User login issue
Hi folks, I guess I messed up something in my ldap settings using GOsa2. Users can't login on the server any longer with the exception of root. This includes logins via console. The first user however can still login into GOsa2 and things look good at first glance. The following also looks weird to me: root@tjener:~# ldapsearch –x uid=roman SASL/GS2-IAKERB authentication started ldap_sasl_interactive_bind_s: Local error (-2) additional info: SASL(-1): generic failure: GS2 Error: Unspecified GSS failure. Minor code may provide more information (No Kerberos credentials available (default cache: FILE:/tmp/krb5cc_0)) root@tjener:~# Any ideas how to fix this? Kind regards, Roman
Re: Problems changing parameters using GOsa2
Hi Mike, > Make sure you click Ok/Apply/Save on all subpages until you are back > in the object list ( e.g. list if systems, users, groups). > > Does that help? Wow, that's embarrassing! Thanks, my bad! Kind regards, Roman
Problems changing parameters using GOsa2
Hi folks, I'm not very familiar with GOsa2. So far I mostly used commandline to edit my ldap settings. I'm facing some rather weird issue when using GOsa2. It appears that the rights of the first user are somehow compromised in my setup. Edits are apparently accepted but when I click a different menu item and then go back to the menu where I did the changes, all changed parameters have been reset and my changes are lost. What am I missing? Kind regards, Roman
Re: Running arm64 and i386 chroot in parallel on the same subnet
Hi Mike, > Running both client types on the same subnet should be possible. Wow, nice! > Create two different DHCP groups and have different boot options > (those you need for i386 / arm64) for each group. > > Then you can assign the arm64 clients to one DHCP group and the i386 > clients to the other. Both on the same subnet. How do I assign clients to a specific DHCP group? > More elegant would be if you could derive the architecture from the > DHCP request somehow. I am not sure if that is possible, but it would > be nice to have. Yep, I don't have a clue as well. Anybody? > Thanks for reporting back here on your success status! Yeah, I'm rather happy how this project is going. School computer labs can become significantly more affordable using Raspberry Pi's. I wouldn't have come so far without your support. Thanks! > Greets, > Mike Kind regards, Roman
Running arm64 and i386 chroot in parallel on the same subnet
Hi folks, I'm able to run both chroot images, arm64 (rpi4) and i386 on the debian edu server. The arm64-one runs inside of a qemu vm. Various adjustments to the configuration of Debian Edu are required before it is possible to run either the one or the other image. However, my goal is to serve and run both images at the same time. I believe I can make this work using two different subnets. However, my preferred approach is, to use the same subnet for both. Any thoughts on this? Kind regards, Roman
Missing epoptes administration tool in clients (diskless workstation)
Hi folks, Using "thin client" I get access to the epoptes administration tool on all clients allowing me to use any client as an epoptes administrator. On "diskless workstations" I'm forced to use the server for administration, which is not convenient in our setup since the server is located in the server room. Is it not possible to have the epoptes administration tool to be installed on our "diskless workstation image" as well or will this break something? How is this done? Kind regards, Roman
How to add comments to ldap?
Hi folks, I usually edit my ldap configuration using "ldapvi -ZD '(cn=admin)'". I was wondering if there is a way to add comments? Especially during tests/debuggings this would be helpful. Kind regards, Roman
Re: Fwd: Re: arm64 images to boot raspberry pi's
Hi folks, Seems I was able to debug this. The problem was related to adjustments to the settings of isc-dhcp (ldap). The following settings worked for me. I have added the default settings which are the commented lines. The qemu arm64 server is connected using a network bridge and runs on 10.0.2.10. 81 cn=intern,cn=dhcp,cn=tjener,ou=servers,ou=systems,dc=skole,dc=skolelinux,dc=no cn: intern objectClass: top objectClass: dhcpSharedNetwork dhcpOption: broadcast-address 10.255.255.255 dhcpOption: domain-name "intern" dhcpOption: domain-name-servers 10.0.2.2 dhcpOption: netbios-name-servers tjener dhcpOption: netbios-node-type 8 dhcpOption: ntp-servers ntp #dhcpOption: root-path "/opt/ltsp/i386" dhcpOption: root-path "/srv/ltsp/raspios" dhcpOption: routers gateway dhcpOption: smtp-server postoffice dhcpOption: subnet-mask 255.0.0.0 dhcpOption: wpad-url "http://wpad/wpad.dat; dhcpStatements: allow booting dhcpStatements: allow bootp dhcpStatements: allow unknown-clients dhcpStatements: authoritative dhcpStatements: default-lease-time 1800 #dhcpStatements: filename "/var/lib/tftpboot/pxelinux.0" dhcpStatements: filename "cmdline.txt" dhcpStatements: max-lease-time 3600 #dhcpStatements: next-server tjener dhcpStatements: next-server 10.0.2.10 Kind regards, Roman > On 12/22/2023 10:59 AM GMT roman.me...@gismap.ch wrote: > > > Hi folks, > > I was able to make my rpi4 boot for the first time today. > > Anyway, I'm getting the following output, which is rather gibberish to me > unfortunately. > > Any ideas what is happening? > > [ 124.894512] VFS: Unable to mount root fs via NFS. > [ 124.910683] Kernel panic - not syncing: Requested init > /usr/share/ltsp/client/inft/init failed (error -2). > [ 124.910789] CPU: 2 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rpi7-rpi/v8 #1 > Debian 1:6.1,63-1+rpt1 > [ 124.910873] Hardware name: Raspberry Pi 4 Model B Rev 1.1 (DT) > [ 124.910929] Call trace: > [ 124.910959] dump_backtrace.part.0+0xec/Ox100 > [ 124.911919] show_stack+0x20/0x30 > [ 124.911060] dump_stack_lvl+0x88/0xb4 > [ 124.911141] dump_stack+0x18/0x34 > [ 124.91115Z] panic+Ox1a0/0x370 > [ 124.911196] kernel_init+Oxec/Ox140 > [ 124,911242] ret_from_fork+Ox10/0x20 > [ 124.911290] SMP: stopping secondary CPUs > [ 124.911338] Kernel Offset: Ox115a00 from Oxffc00800 > [ 124.911392] PHYS_OFFSET: 0x0 > [ 124.911426] CPU features: 0xB,2013c080,421b > [ 124.911476] Memory Limit: none > [ 124.9115163 ---[ end Kernel panic - not syncing: Requested init > /usr/share/ltsp/client/init/init failed (error -2). ]--- > > Kind regards, > Roman > > > -- Original Message -- > > From: roman.me...@gismap.ch > > To: Mike Gabriel > > Cc: Petter Reinholdtsen , debian-edu@lists.debian.org > > Date: 12/14/2023 1:30 PM GMT > > Subject: Re: arm64 images to boot raspberry pi's > > > > > > Hi Mike, > > > > I was able to add the arm64 LTSP server in a qemu-arm64 VM and also created > > the raspberry pi client image. > > > > I basically followed the instructions on ltsp.org to do so. > > > > The challenge is now to adjust the configuration of Debian Edu and > > qemu-arm64 VM. > > > > Currently the original one on Debian Edu is working just fine and serving > > images to all i386 devices, both diskless and thin client. > > > > The new arm64-ltsp server is used to serve diskless images for the rpi4 but > > is still untested since adjustments to the configuration are needed. > > > > Is there a way to run both ltsp servers in parallel on the same subnet? > > > > Do I have to run each on its specific subnet, e.g. by adding a third > > network card? > > > > What adjustments in the configuration are needed to be able to switch from > > one ltsp server to the other. > > > > Kind regards, > > Roman > > > > > On 12/04/2023 9:01 PM GMT Mike Gabriel > > > wrote: > > > > > > > > > Hi Roman, > > > > > > On Mo 04 Dez 2023 11:53:46 CET, roman.meier wrote: > > > > > > > Hi Petter! > > > > > > > > Thanks for the link! > > > > > > > > I may not have been clear enough. Sorry for that. > > > > > > > > I want to build an image for arm64 to be able to run the raspberry > > > > pi's as ltsp clients. Since newer raspberry pi's are bootnet capable > > > > and raspberry os is based on debian I thought there might be a good > > > > chance for this to work out just fine. > > > > > > > > Kind regards, > > > > Roman > > > > > > We don't have a cross-build approach for LTSP images, yet. What you could > > > do: > > > > > >* install an arm64 LTSP server in a qemu-arm64 VM and create your LTSP > > > client image in there > > >* same on a Rasberry Pi > > >* provide a patch to debian-edu-ltsp-install so that > > > qemu-debootstrap gets used > > > to create that image (see: > > > > > > https://github.com/faiproject/fai/blob/master/bin/fai-make-nfsroot#L277) > > > > > > Greets, > > > Mike > > > -- > > > > > > DAS-NETZWERKTEAM > > > c\o
Fwd: Re: arm64 images to boot raspberry pi's
Hi folks, I was able to make my rpi4 boot for the first time today. Anyway, I'm getting the following output, which is rather gibberish to me unfortunately. Any ideas what is happening? [ 124.894512] VFS: Unable to mount root fs via NFS. [ 124.910683] Kernel panic - not syncing: Requested init /usr/share/ltsp/client/inft/init failed (error -2). [ 124.910789] CPU: 2 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rpi7-rpi/v8 #1 Debian 1:6.1,63-1+rpt1 [ 124.910873] Hardware name: Raspberry Pi 4 Model B Rev 1.1 (DT) [ 124.910929] Call trace: [ 124.910959] dump_backtrace.part.0+0xec/Ox100 [ 124.911919] show_stack+0x20/0x30 [ 124.911060] dump_stack_lvl+0x88/0xb4 [ 124.911141] dump_stack+0x18/0x34 [ 124.91115Z] panic+Ox1a0/0x370 [ 124.911196] kernel_init+Oxec/Ox140 [ 124,911242] ret_from_fork+Ox10/0x20 [ 124.911290] SMP: stopping secondary CPUs [ 124.911338] Kernel Offset: Ox115a00 from Oxffc00800 [ 124.911392] PHYS_OFFSET: 0x0 [ 124.911426] CPU features: 0xB,2013c080,421b [ 124.911476] Memory Limit: none [ 124.9115163 ---[ end Kernel panic - not syncing: Requested init /usr/share/ltsp/client/init/init failed (error -2). ]--- Kind regards, Roman > -- Original Message -- > From: roman.me...@gismap.ch > To: Mike Gabriel > Cc: Petter Reinholdtsen , debian-edu@lists.debian.org > Date: 12/14/2023 1:30 PM GMT > Subject: Re: arm64 images to boot raspberry pi's > > > Hi Mike, > > I was able to add the arm64 LTSP server in a qemu-arm64 VM and also created > the raspberry pi client image. > > I basically followed the instructions on ltsp.org to do so. > > The challenge is now to adjust the configuration of Debian Edu and qemu-arm64 > VM. > > Currently the original one on Debian Edu is working just fine and serving > images to all i386 devices, both diskless and thin client. > > The new arm64-ltsp server is used to serve diskless images for the rpi4 but > is still untested since adjustments to the configuration are needed. > > Is there a way to run both ltsp servers in parallel on the same subnet? > > Do I have to run each on its specific subnet, e.g. by adding a third network > card? > > What adjustments in the configuration are needed to be able to switch from > one ltsp server to the other. > > Kind regards, > Roman > > > On 12/04/2023 9:01 PM GMT Mike Gabriel > > wrote: > > > > > > Hi Roman, > > > > On Mo 04 Dez 2023 11:53:46 CET, roman.meier wrote: > > > > > Hi Petter! > > > > > > Thanks for the link! > > > > > > I may not have been clear enough. Sorry for that. > > > > > > I want to build an image for arm64 to be able to run the raspberry > > > pi's as ltsp clients. Since newer raspberry pi's are bootnet capable > > > and raspberry os is based on debian I thought there might be a good > > > chance for this to work out just fine. > > > > > > Kind regards, > > > Roman > > > > We don't have a cross-build approach for LTSP images, yet. What you could > > do: > > > >* install an arm64 LTSP server in a qemu-arm64 VM and create your LTSP > > client image in there > >* same on a Rasberry Pi > >* provide a patch to debian-edu-ltsp-install so that > > qemu-debootstrap gets used > > to create that image (see: > > > > https://github.com/faiproject/fai/blob/master/bin/fai-make-nfsroot#L277) > > > > Greets, > > Mike > > -- > > > > DAS-NETZWERKTEAM > > c\o Technik- und Ökologiezentrum Eckernförde > > Mike Gabriel, Marienthaler Str. 17, 24340 Eckernförde > > mobile: +49 (1520) 1976 148 > > landline: +49 (4351) 850 8940 > > > > GnuPG Fingerprint: 9BFB AEE8 6C0A A5FF BF22 0782 9AF4 6B30 2577 1B31 > > mail: mike.gabr...@das-netzwerkteam.de, http://das-netzwerkteam.de
Re: arm64 images to boot raspberry pi's
Hi Mike, I was able to add the arm64 LTSP server in a qemu-arm64 VM and also created the raspberry pi client image. I basically followed the instructions on ltsp.org to do so. The challenge is now to adjust the configuration of Debian Edu and qemu-arm64 VM. Currently the original one on Debian Edu is working just fine and serving images to all i386 devices, both diskless and thin client. The new arm64-ltsp server is used to serve diskless images for the rpi4 but is still untested since adjustments to the configuration are needed. Is there a way to run both ltsp servers in parallel on the same subnet? Do I have to run each on its specific subnet, e.g. by adding a third network card? What adjustments in the configuration are needed to be able to switch from one ltsp server to the other. Kind regards, Roman > On 12/04/2023 9:01 PM GMT Mike Gabriel > wrote: > > > Hi Roman, > > On Mo 04 Dez 2023 11:53:46 CET, roman.meier wrote: > > > Hi Petter! > > > > Thanks for the link! > > > > I may not have been clear enough. Sorry for that. > > > > I want to build an image for arm64 to be able to run the raspberry > > pi's as ltsp clients. Since newer raspberry pi's are bootnet capable > > and raspberry os is based on debian I thought there might be a good > > chance for this to work out just fine. > > > > Kind regards, > > Roman > > We don't have a cross-build approach for LTSP images, yet. What you could do: > >* install an arm64 LTSP server in a qemu-arm64 VM and create your LTSP > client image in there >* same on a Rasberry Pi >* provide a patch to debian-edu-ltsp-install so that > qemu-debootstrap gets used > to create that image (see: > https://github.com/faiproject/fai/blob/master/bin/fai-make-nfsroot#L277) > > Greets, > Mike > -- > > DAS-NETZWERKTEAM > c\o Technik- und Ökologiezentrum Eckernförde > Mike Gabriel, Marienthaler Str. 17, 24340 Eckernförde > mobile: +49 (1520) 1976 148 > landline: +49 (4351) 850 8940 > > GnuPG Fingerprint: 9BFB AEE8 6C0A A5FF BF22 0782 9AF4 6B30 2577 1B31 > mail: mike.gabr...@das-netzwerkteam.de, http://das-netzwerkteam.de
Re: Keyboard and mouse not working on lenovo laptop
Hi folks, I can't really explain this, but after rebuilding the ltsp images, things worked just fine. Kind regards, Roman > On 12/02/2023 8:28 AM GMT roman.me...@gismap.ch wrote: > > > Hi folks, > > One of our teachers tried to boot his laptop for the first time directly from > Debian Edu. The debian edu boot screen is nicely showing but neither thin > client nor diskless workstation works for him. He gets to the login screen > though but neither keyboard nor mouse are working and therefore he is not > able to login. > > Any ideas how to fix this? > > Kind regards, > Roman
Re: arm64 images to boot raspberry pi's
Hi Petter! Thanks for the link! I may not have been clear enough. Sorry for that. I want to build an image for arm64 to be able to run the raspberry pi's as ltsp clients. Since newer raspberry pi's are bootnet capable and raspberry os is based on debian I thought there might be a good chance for this to work out just fine. Kind regards, Roman > On 12/03/2023 12:18 PM GMT Petter Reinholdtsen wrote: > > > [Roman Meier] > > AFAIK debian edu currently only supports amd64 and i386. > > > > Is it possible to add arm64 to be used e.g. with raspberry pi's? > > You might find something useful on > https://wiki.debian.org/DebianEdu/HowTo/RaspberryPiWorkstation >. > > -- > Happy hacking > Petter Reinholdtsen
arm64 images to boot raspberry pi's
Hi folks, AFAIK debian edu currently only supports amd64 and i386. Is it possible to add arm64 to be used e.g. with raspberry pi's? Kind regards, Roman
Keyboard and mouse not working on lenovo laptop
Hi folks, One of our teachers tried to boot his laptop for the first time directly from Debian Edu. The debian edu boot screen is nicely showing but neither thin client nor diskless workstation works for him. He gets to the login screen though but neither keyboard nor mouse are working and therefore he is not able to login. Any ideas how to fix this? Kind regards, Roman
Re: Transfer storage from home to var
Hi Dashamir, The reassigning seemed to finally work using: sudo lvresize -r -L-10G /dev/vg_system/skole+tjener+home0 sudo lvresize -r -L+10G /dev/vg_system/var Then I'm getting the following output: root@tjener:/skole/tjener/home0/roman# sudo vgdisplay --- Volume group --- VG Name vg_system System ID Formatlvm2 Metadata Areas1 Metadata Sequence No 79 VG Access read/write VG Status resizable MAX LV0 Cur LV10 Open LV 10 Max PV0 Cur PV1 Act PV1 VG Size 930.55 GiB PE Size 4.00 MiB Total PE 238222 Alloc PE / Size 238088 / 930.03 GiB Free PE / Size 134 / 536.00 MiB VG UUID 1Ekw6T-kY0s-iJjc-HV4x-ldxG-kQNh-ZTAATE root@tjener:/skole/tjener/home0/roman# sudo lvdisplay --- Logical volume --- LV Path/dev/vg_system/root LV Nameroot VG Namevg_system LV UUIDKm3Z2m-ei4I-zPaG-AanS-oPsF-4y1u-BmGGag LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:48 + LV Status available # open 1 LV Size<38.36 GiB Current LE 9820 Segments 7 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:0 --- Logical volume --- LV Path/dev/vg_system/usr LV Nameusr VG Namevg_system LV UUIDyhlCZE-cqhZ-DvAd-aHBc-Mdhr-nwDS-abfPMN LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:49 + LV Status available # open 1 LV Size25.00 GiB Current LE 6400 Segments 4 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:2 --- Logical volume --- LV Path/dev/vg_system/var LV Namevar VG Namevg_system LV UUIDFcNFe9-yd08-DKtF-u3Cl-OGvr-mF48-USUnAV LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:49 + LV Status available # open 1 LV Size<27.26 GiB Current LE 6978 Segments 5 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:3 --- Logical volume --- LV Path/dev/vg_system/var+log LV Namevar+log VG Namevg_system LV UUIDMYTIao-Vio4-jRik-0gb3-Egg7-sQbc-UMaeNe LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:49 + LV Status available # open 1 LV Size<4.77 GiB Current LE 1220 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:4 --- Logical volume --- LV Path/dev/vg_system/swap_1 LV Nameswap_1 VG Namevg_system LV UUIDlQSvdd-hcUB-9bdV-LkN1-UiMI-PhqC-zLW71E LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:49 + LV Status available # open 2 LV Size3.81 GiB Current LE 976 Segments 1 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:1 --- Logical volume --- LV Path/dev/vg_system/opt LV Nameopt VG Namevg_system LV UUIDQ23PB4-n1bk-H3Xg-kjjO-NbUA-AZ8n-9TWQzw LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:50 + LV Status available # open 1 LV Size<37.52 GiB Current LE 9604 Segments 3 Allocation inherit Read ahead sectors auto - currently set to 256 Block device 254:5 --- Logical volume --- LV Path/dev/vg_system/var+opt+ltsp+swapfiles LV Namevar+opt+ltsp+swapfiles VG Namevg_system LV UUIDjpN2ka-TLP3-1GsX-lUlV-Khz7-7QGt-BkM8km LV Write Accessread/write LV Creation host, time localhost, 2019-12-22 00:17:50 + LV Status available # open 1 LV Size3.81 GiB Current LE 976 Segments 1 Allocation inherit Read ahead sectors auto - currently set to
Re: Transfer storage from home to var
Hi Petter, The following worked based on your suggestion: sudo lvresize -r -L-10G /dev/vg_system/skole+tjener+home0 sudo lvresize -r -L+10G /dev/vg_system/var I hope this was safe and nothing got lost on home0 ;) Kind regards, Roman > On 02/03/2023 9:04 AM GMT Petter Reinholdtsen wrote: > > > [Roman Meier] > > However, extending the size of /var didn't work: > > > > sudo lvextend -L+5G /dev/vg_system/var > > Insufficient free space: 1280 extents needed, but only 138 available > > I would strongly recommend not using lvextend, and instead use lvresize > -r. I suspect the underlying LVM logical volume is already increased, > but the file system on top is not, and thus you do not see the space. > > Perhaps 'lvresize -r -L+1M /dev/vg_system/var' work better. > > -- > Happy hacking > Petter Reinholdtsen
Re: Transfer storage from home to var
Hi Petter, I wrote the email a bit too hastily. Sorry for causing this confusion. It's /var and not /dev. Again, I was able to nicely reduce the size of /skole/jener/home0. However, extending the size of /var didn't work: sudo lvextend -L+5G /dev/vg_system/var Insufficient free space: 1280 extents needed, but only 138 available Kind regards, Roman > On 02/02/2023 9:21 PM GMT Petter Reinholdtsen wrote: > > > [Roman Meier] > > As mentioned in my email, I was able to reduce the size of > > /skole/jener/home0. I therefore logged in als root and unmounted > > /skole/jener/home0. > > OK, then I misunderstood. I read your email to say that you chould not > find a way to reduce the size. > > > For unknown reasons this space is however not made available to be > > reassigned to /dev resulting in the error message mentioned. > > Reducing home in LVM would just make ununsed space available to LVM, > visible using vgs. This unused space chan then be allocated to other > partitions using lvresize on this partition. > > You might want to check if you really need space on /dev/, it seem like > the wrong partition to extend. > -- > Happy hacking > Petter Reinholdtsen
Re: Transfer storage from home to var
Hi Petter, Thanks for your fast reply! As mentioned in my email, I was able to reduce the size of /skole/jener/home0. I therefore logged in als root and unmounted /skole/jener/home0. For unknown reasons this space is however not made available to be reassigned to /dev resulting in the error message mentioned. Kind regards, Roman > On 02/02/2023 12:23 PM GMT Petter Reinholdtsen wrote: > > > [Roman Meier] > > I have run out of storage on /var and I can't extend size because all > > on the harddrive has been used already. > > > > I have about 50G unused space on /skole/jener/home0 but I find it > > difficult to transfer storage from /skole/jener/home to /dev. > > I normally boot in single user, unmount /skole/jener/home0 and run > lvresize -r -L-10g /dev/vg_system/home or similar. > > -- > Happy hacking > Petter Reinholdtsen
Transfer storage from home to var
Hi folks, I have run out of storage on /var and I can't extend size because all on the harddrive has been used already. I have about 50G unused space on /skole/jener/home0 but I find it difficult to transfer storage from /skole/jener/home to /dev. I was able to reduce storage of /skole/jener/home by about 10G but assigning 5G of it to /dev results in an error message: root@tjener:/skole/tjener/home0/roman# sudo lvextend -L+5G /dev/vg_system/var Insufficient free space: 1280 extents needed, but only 138 available root@tjener:/skole/tjener/home0/roman# sudo lvextend -L+9G /dev/vg_system/var Any ideas? Kind regards, Roman
First user
Hi folks, The first user changed his password and is saying, he can't log in anymore. How can I fix this? Kind regards, Roman
Remote desktop client for Ubuntu
Hi folks, Some of our teachers prefer to use Windows on their laptops. They use a "remote desktop connection" to use Debian Edu and it works nicely for them. We do have teachers using Linux aswell, e.g. Ubuntu. What is the preferred tool to use Debian Edu for them? Kind regards, Roman
Re: Accessing the backup system (slbackup-php)
Yep, that helped. Thanks Wolfgang. :) > On 05/27/2021 7:43 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-05-27 ] > > I noticed that I can access Debian Edu's backup system (slbackup-php) > > only if I log in on the server. If I'm trying to access it e.g. on my > > workstation or a ltsp client, I don't get the login screen and I'm > > redirected to a forbidden page. > > Maybe using IP address 10.0.2.2 (default) instead of 'www' would help. > > Wolfgang
Accessing the backup system (slbackup-php)
Hi folks, I noticed that I can access Debian Edu's backup system (slbackup-php) only if I log in on the server. If I'm trying to access it e.g. on my workstation or a ltsp client, I don't get the login screen and I'm redirected to a forbidden page. Is this normal behaviour, maybe for security reasons? Is it possible to enable other devices like my workstation to have access to the backup menu? Kind regards, Roman
Incomplete bulk user import
Hi folks, I tried to import users via CSV Import this morning. My list contains 324 students. When executing the import the browser shows a blank page after a few minutes without throwing the usual summary. I then waited some time to be sure the process is done. However the number of created users is 289 instead of 324. How can I ensure that the summary page is showing for bigger imports aswell? It works just fine for the teachers with a count of just 13. Kind regards, Roman
How to setup a browser extension for all users by default?
Hi folks, I have a browser extension that I want to setup by default for all users and for both browsers, Chromium and Firefox. I want to avoid that every user needs to do this manually. Can this be done and how? Kind regards, Roman
Re: Problems installing Epoptes for "diskless workstation"
Hi Wolfgang, I did more tests today with Epoptes on "diskless workstation". Service epoptes-client fails to start upon booting. The reason seems to be that the server certificate file is missing. I can manually run the following command on the client via console to fetch the certificate: /usr/sbin/epoptes-client -c Then I can start the epoptes-client service and Epoptes works fine. I think the above should be done upon booting, right? How is this done? Another question: The Epoptes admin gui is missing on "diskless workstation". Therefore a teacher can't use Epoptes to assist the kids unless he switches to "thin client". How can this be added for "diskless workstation" as well? Kind regards, Roman > On 04/24/2021 7:29 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-24 ] > > https://www/debian-edu-doc/en/debian-edu-buster-manual.html#TeachAndLearn--Monitoring_pupils > > > > The following line doesn't work any longer because > > /etc/init.d/epoptes-client can't be found: > > > > ltsp-chroot -m --arch amd64 sed -i 's/test -f/#test -f/' > > /etc/init.d/epoptes-client > > > > Any ideas, what's wrong? > > The documentation is outdated. Epoptes switched from init file to > systemd service file. > > I'll update the manual. > > Thanks for noticing, > > Wolfgang
Re: Running Kiwix-Server on Debian Edu
Hi Petter, Yep, both http://www:89/ and http://tjener:89/ work just fine! Thanks a lot! Kind regards, Roman > On 04/29/2021 10:23 AM Petter Reinholdtsen wrote: > > > [Roman Meier] > > How can I fix this? > > I suspect it might work better if you use the hostname instead of the IP > address of the kiwix server. Did you try this? > > -- > Happy hacking > Petter Reinholdtsen
Running Kiwix-Server on Debian Edu
Hi folks, I would like to run Kiwix on Debian Edu. Kiwix provides an offline version of Wikipedia. I can launch the server using the following command: ./kiwix-serve --port=89 --verbose /etc/debian-edu/www/wiki/wiktionary_en_all_maxi_2021-02.zim When a user then tries to access Wikipedia with his browser using http://10.0.2.2:89/ Squid blocks access throwing "Access Denied" unless I stop the squid server. How can I fix this? Kind regards, Roman
Re: Client time/date synchronisation
Hi Wolfgang, > On 04/27/2021 11:36 AM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-27 ] > > However, I noticed that the value of FAT_RAM_THRESHOLD is ignored. All > > devices whether they have more or less than 1000KB of RAM boot as > > "thin client". Somehow it's acting like setting > > LTSP_FATCLIENT="False". The threshold value was working well for me in > > earlier installations of Debian Edu. > > Please check if you still have LTSP_FATCLIENT="False" in LDAP (or > somewhere else). I don't think so. I triple checked that one. > > Also, rebuilding the image seems to have reduced the content of the > > menu tree appearing when clicking Applications. It now starts with > > "Settings" and ends with "System". > > Strange. Any other changes done? > Does this also happen if using a fresh account? Ok, I've created a new account and this one is showing the complete menu tree. How come and how to I fix the old ones? > Wolfgang Kind regards, Roman
How to force "thin client" in the secondary subnet only
Hi folks, I'm currently forcing "thin client" using LTSP_FATCLIENT="False". This works nicely for all devices of our computer lab connected to the secondary subnet. Unfortunately this also affects devices connected to the primary subnet. The PXE menu still offers "diskless workstation" but because "thin client" is forced it's not booting as "diskless workstation" anymore. I would like to offer all users connected to the primary subnet a choice to boot as "diskless workstation" or "thin client" and not being forced into "thin client" as in the secondary subnet. How can I achieve this? Kind regards, Roman
Re: Client time/date synchronisation
Hi Wolfgang, Yep, this worked, thanks! However, I noticed that the value of FAT_RAM_THRESHOLD is ignored. All devices whether they have more or less than 1000KB of RAM boot as "thin client". Somehow it's acting like setting LTSP_FATCLIENT="False". The threshold value was working well for me in earlier installations of Debian Edu. Also, rebuilding the image seems to have reduced the content of the menu tree appearing when clicking Applications. It now starts with "Settings" and ends with "System". The complete menu tree however appears nicely when I do a right click on the desktop> Applications also including, e.g. "Run Program...", "Log Out", etc. Kind regards, Roman > On 04/24/2021 7:21 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-24 ] > > I still have an issue with "FAT_RAM_THRESHOLD=1000" in > > "/var/lib/tftpboot/ltsp/i386/lts.conf". > > No idea why it doesn't take effect. > > Edit /opt/ltsp/i386/etc/lts.conf instead and run 'ltsp-update-image' > before starting an LTSP client. > > Wolfgang
Problems installing Epoptes for "diskless workstation"
Hi folks, Epoptes is working just fine using "thin client" for me. I'm now trying to make it work on "diskless workstation" aswell. I'm using the guide found here: https://www/debian-edu-doc/en/debian-edu-buster-manual.html#TeachAndLearn--Monitoring_pupils The following line doesn't work any longer because /etc/init.d/epoptes-client can't be found: ltsp-chroot -m --arch amd64 sed -i 's/test -f/#test -f/' /etc/init.d/epoptes-client Any ideas, what's wrong? Kind regards, Roman
Re: Epoptes broadcasting using thin client
Hi folks, Let me answer the question myself. It seems that a packages was missing in my installation. "apt install ssvnc" fixed it. Both fullscreen and windowed broadcast started to work just fine after rebooting the thin clients. Kind regards, Roman > On 04/09/2021 3:51 PM roman.me...@gismap.ch wrote: > > > Hi folks, > > I'm currently using thin client only. Assisting the students using Epoptes > works nicely. > > However, anytime I try to use either the fullscreen or windowed broadcast, > nothing shows on any of the selected clients. > > Was anyone able to make Epoptes broadcasting work on Debian Edu? > > Any hints why this is happening and how I can fix it? > > Kind regards, > Roman
Fwd: Re: Client time/date synchronisation
Hi Wolfgang, I still have an issue with "FAT_RAM_THRESHOLD=1000" in "/var/lib/tftpboot/ltsp/i386/lts.conf". I noticed that this only works in the "primary subnet". Booting via "secondary subnet" always starts the device as "diskless workstation". This is weird to me and I don't think I have seen this before. Any ideas why? Kind regards, Roman > -- Original Message -- > From: roman.me...@gismap.ch > To: Wolfgang Schweer , debian-edu@lists.debian.org > Date: 04/23/2021 2:53 PM > Subject: Re: Client time/date synchronisation > > > Hi Wolfgang, > > I have done a new installation this morning using the i386-server-image > instead of the amd64 one. Things definitely work much better now. > > Client time sync without internet was not working for me with the default > setting "ltspConfig: TIMESERVER=ntp". You suggested to switch to "ltspConfig: > TIMESERVER=auto". This works for me appending "server 10.0.2.2" to > /etc/ntp.conf. Syncing however happens with several minutes delay. > > Thin client is also working using "ltspConfig: LTSP_FATCLIENT="False"". > However, this setting is not what I prefer because it deactivates the > possibility to boot as "diskless workstation" in the "primary subnet". > > Removing "ltspConfig: LTSP_FATCLIENT="False"" and adding "ltspConfig: > FAT_RAM_THRESHOLD=1000" should work but doesn't for me. I don't know why. > > The fix for me was to remove "ltspConfig: LTSP_FATCLIENT="False"" again and > editing the configuration in /var/lib/tftpboot/ltsp/i386/lts.conf by adding > "FAT_RAM_THRESHOLD=1000". > > Thanks a lot for your help fixing my issues. > > Kind regards, > Roman > > > On 04/22/2021 2:52 PM Wolfgang Schweer wrote: > > > > > > Hi Roman, > > > > [ roman.me...@gismap.ch, 2021-04-22 ] > > > > > This morning I decided to switch back to the original amd64 image and > > > testing with a newer client device I just got. > > > > > > The good news is that everything seems to work just fine. The device > > > is booting nicely as thin client. LDAP settings are added as expected > > > to /run/ltsp/ltsp_config_env and there's a new entry "server ntp" at > > > the end of /etc/ntp.conf. > > > > Good. > > > > > These things are not happening when I boot using the i386 image. > > > More ideas why this is happening? > > > > Might be due to the 64-bit/32-bit mix. I don't have a test environment > > to check it, though. > > > > As another option to get the Toughbooks working, you could install the > > combined server using the 32-bit BD ISO image, see the status page: > > https://wiki.debian.org/DebianEdu/Status/Buster#Download_using_http > > > > The LTSP chroot will be 32-bit (i386) by default in this case (so no > > $arch mix) and things should work also for 64-bit clients. > > > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, I tried using iburst as you suggested. It seems not to sync anymore. Anyway, I'll do more tests and if I find something I'll post my findings. Kind regards, Roman > On 04/23/2021 3:31 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-23 ] > > I have done a new installation this morning using the > > i386-server-image instead of the amd64 one. Things definitely work > > much better now. > > Good to hear. > > > Client time sync without internet was not working for me with the > > default setting "ltspConfig: TIMESERVER=ntp". You suggested to switch > > to "ltspConfig: TIMESERVER=auto". This works for me appending "server > > 10.0.2.2" to /etc/ntp.conf. Syncing however happens with several > > minutes delay. > > 10.0.2.2 isn't a public server, so iburst as an 'aggressive' option > could be used to speed up things (man ntp.conf): > > ltspConfig: TIMESERVER="10.0.2.2 iburst" > > Maybe it works like expected… > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, I have done a new installation this morning using the i386-server-image instead of the amd64 one. Things definitely work much better now. Client time sync without internet was not working for me with the default setting "ltspConfig: TIMESERVER=ntp". You suggested to switch to "ltspConfig: TIMESERVER=auto". This works for me appending "server 10.0.2.2" to /etc/ntp.conf. Syncing however happens with several minutes delay. Thin client is also working using "ltspConfig: LTSP_FATCLIENT="False"". However, this setting is not what I prefer because it deactivates the possibility to boot as "diskless workstation" in the "primary subnet". Removing "ltspConfig: LTSP_FATCLIENT="False"" and adding "ltspConfig: FAT_RAM_THRESHOLD=1000" should work but doesn't for me. I don't know why. The fix for me was to remove "ltspConfig: LTSP_FATCLIENT="False"" again and editing the configuration in /var/lib/tftpboot/ltsp/i386/lts.conf by adding "FAT_RAM_THRESHOLD=1000". Thanks a lot for your help fixing my issues. Kind regards, Roman > On 04/22/2021 2:52 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-22 ] > > > This morning I decided to switch back to the original amd64 image and > > testing with a newer client device I just got. > > > > The good news is that everything seems to work just fine. The device > > is booting nicely as thin client. LDAP settings are added as expected > > to /run/ltsp/ltsp_config_env and there's a new entry "server ntp" at > > the end of /etc/ntp.conf. > > Good. > > > These things are not happening when I boot using the i386 image. > > More ideas why this is happening? > > Might be due to the 64-bit/32-bit mix. I don't have a test environment > to check it, though. > > As another option to get the Toughbooks working, you could install the > combined server using the 32-bit BD ISO image, see the status page: > https://wiki.debian.org/DebianEdu/Status/Buster#Download_using_http > > The LTSP chroot will be 32-bit (i386) by default in this case (so no > $arch mix) and things should work also for 64-bit clients. > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, I have new findings which may help to narrow things down hopefully. This morning I decided to switch back to the original amd64 image and testing with a newer client device I just got. The good news is that everything seems to work just fine. The device is booting nicely as thin client. LDAP settings are added as expected to /run/ltsp/ltsp_config_env and there's a new entry "server ntp" at the end of /etc/ntp.conf. These things are not happending when I boot using the i386 image. More ideas why this is happening? Kind regards, Roman > On 04/21/2021 4:32 PM Wolfgang Schweer wrote: > > > [ roman.me...@gismap.ch, 2021-04-21 ] > > I have added the suggested line to LDAP. It looks like this now: > > > > 121 cn=ltspConfigDefault,ou=ltsp,dc=skole,dc=skolelinux,dc=no > > objectClass: ltspClientConfig > > cn: ltspConfigDefault > > ltspConfig: KEEP_SYSTEM_SERVICES=lightdm > > ltspConfig: LTSP_FATCLIENT="False" > > ltspConfig: TIMESERVER=ntp > > > > After this, my client still boots as "diskless workstation" and the > > content of /run/ltsp/ltsp_config_env shows LTSP_FATCLIENT="True" > > instead of "False". > > That's weird, but I expected something like this because ntp.conf was > left unchanged according to your report. > > There should be LTSP_FATCLIENT="True" in ltsp_config_env before the > three mentioned settings, but LTSP_FATCLIENT="False" at the end > (overriding the first occurence). > > > Are changes to LDAP validated on-the-fly upon save or do I need to add > > another step like commit, rebuilding the image, etc.? > > No additional step needed. During LTSP client boot, some scripts are > executed to configure the client (overlay filesystem). The scripts are > located in the /usr/share/ltsp/init-ltsp.d/ directory inside the > SquashFS image resp. in /opt/ltsp/i386/usr/share/ltsp/init-ltsp.d/ > > Please check those places. > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, Sorry, still no good news. I have added the suggested line to LDAP. It looks like this now: 121 cn=ltspConfigDefault,ou=ltsp,dc=skole,dc=skolelinux,dc=no objectClass: ltspClientConfig cn: ltspConfigDefault ltspConfig: KEEP_SYSTEM_SERVICES=lightdm ltspConfig: LTSP_FATCLIENT="False" ltspConfig: TIMESERVER=ntp After this, my client still boots as "diskless workstation" and the content of /run/ltsp/ltsp_config_env shows LTSP_FATCLIENT="True" instead of "False". Are changes to LDAP validated on-the-fly upon save or do I need to add another step like commit, rebuilding the image, etc.? Kind regards, Roman > On 04/21/2021 1:31 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-21 ] > > I also played around with LDAP unsuccessfully trying to increasing the > > threshold for my clients to boot as thin client. > > Maybe it doesn't take effect anymore for some reason… > But you should be able to do this by setting the client type in LDAP > > ltspConfig: LTSP_FATCLIENT="False" > > You should be able to see the environment variables actually used: > cat /run/ltsp/ltsp_config_env (diskless workstation) > > If running a thin client, log in graphically as first user. Then switch > to a terminal (Ctrl-Alt-F1), log in as first user, run 'sudo su -' to > become root and run 'cat /run/ltsp/ltsp_config_env' and maybe check > other things, too. > > > That's all I can remember. Is the log file I sent not helpful? > > I can't find anything suspicious… > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, Well, I had to create the i386 image myself because it is not created by default any longer. I also played around with LDAP unsuccessfully trying to increasing the threshold for my clients to boot as thin client. That's all I can remember. Is the log file I sent not helpful? Kind regards, Roman > On 04/21/2021 11:40 AM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-21 ] > > > I just checked the content of /etc/ntp.conf of my "diskless > > workstation" client. No line "server ntp" has been appended. > > > > Guess that explains the problem I'm having, right? > > Yes; any LDAP or LTSP chroot changes on your side? > > Wolfgang
Fwd: Re: Client time/date synchronisation
Hi Wolfgang, Let me also add the syslog. It may probably help: Apr 21 08:51:18 tjener dhcpd[1627]: reuse_lease: lease age 93385 (secs) under 25% threshold, reply with unaltered, existing lease for 192.168.0.20 Apr 21 08:51:18 tjener dhcpd[1627]: DHCPDISCOVER from 00:d0:59:22:05:43 via eth1 Apr 21 08:51:18 tjener dhcpd[1627]: DHCPOFFER on 192.168.0.20 to 00:d0:59:22:05:43 via eth1 Apr 21 08:51:18 tjener dhcpd[1627]: reuse_lease: lease age 93385 (secs) under 25% threshold, reply with unaltered, existing lease for 192.168.0.20 Apr 21 08:51:18 tjener dhcpd[1627]: DHCPREQUEST for 192.168.0.20 (192.168.0.254) from 00:d0:59:22:05:43 via eth1 Apr 21 08:51:18 tjener dhcpd[1627]: DHCPACK on 192.168.0.20 to 00:d0:59:22:05:43 via eth1 Apr 21 08:51:18 tjener nbd_server[1121]: Spawned a child process Apr 21 08:51:18 tjener nbd_server[14337]: virtstyle ipliteral Apr 21 08:51:18 tjener nbd_server[14337]: connect from 192.168.0.20, assigned file is /opt/ltsp/images/i386.img Apr 21 08:51:18 tjener nbd_server[14337]: Can't open authorization file /etc/ltsp/nbd-server.allow (No such file or directory). Apr 21 08:51:18 tjener nbd_server[14337]: Size of exported file/device is 3733676032 Apr 21 08:51:18 tjener nbd_server[14337]: Starting to serve Apr 21 08:51:23 tjener named[1016]: resolver priming query complete Apr 21 08:51:23 tjener ntpd[1029]: error resolving pool 2.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:51:33 tjener named[1016]: resolver priming query complete Apr 21 08:51:33 tjener ntpd[1029]: error resolving pool 0.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:51:43 tjener named[1016]: resolver priming query complete Apr 21 08:51:43 tjener ntpd[1029]: error resolving pool 3.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:51:53 tjener named[1016]: resolver priming query complete Apr 21 08:51:53 tjener ntpd[1029]: error resolving pool 1.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:52:27 tjener nbd_server[1121]: Spawned a child process Apr 21 08:52:27 tjener nbd_server[14491]: virtstyle ipliteral Apr 21 08:52:27 tjener nbd_server[14491]: connect from 192.168.0.20, assigned file is /var/opt/ltsp/swapfiles/192.168.0.20 Apr 21 08:52:27 tjener nbd_server[14491]: Can't open authorization file (No such file or directory). Apr 21 08:52:27 tjener nbd_server[14491]: Client '192.168.0.20' not allowed access by prerun script Apr 21 08:52:27 tjener nbd_server[14491]: Negotiation failed: Connection reset by peer Apr 21 08:52:27 tjener nbd_server[14491]: Exiting. Apr 21 08:52:27 tjener nbd_server[1121]: Child exited with 1 Apr 21 08:52:30 tjener named[1016]: resolver priming query complete Apr 21 08:52:30 tjener ntpd[1029]: error resolving pool 2.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:52:40 tjener named[1016]: resolver priming query complete Apr 21 08:52:40 tjener ntpd[1029]: error resolving pool 0.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:52:50 tjener ntpd[1029]: error resolving pool 3.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:52:50 tjener named[1016]: resolver priming query complete Apr 21 08:53:00 tjener named[1016]: resolver priming query complete Apr 21 08:53:00 tjener ntpd[1029]: error resolving pool 1.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:53:19 tjener slapd[1247]: get_filter: conn 1151 unknown attribute type=wWWHomePage (17) Apr 21 08:53:19 tjener slapd[1247]: get_filter: conn 1152 unknown attribute type=wWWHomePage (17) Apr 21 08:53:20 tjener named[1016]: resolver priming query complete Apr 21 08:53:30 tjener named[1016]: resolver priming query complete Apr 21 08:53:35 tjener ntpd[1029]: error resolving pool 2.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:53:45 tjener named[1016]: resolver priming query complete Apr 21 08:53:45 tjener ntpd[1029]: error resolving pool 3.debian.pool.ntp.org: Name or service not known (-2) Apr 21 08:53:55 tjener named[1016]: resolver priming query complete Apr 21 08:53:55 tjener ntpd[1029]: error resolving pool 0.debian.pool.ntp.org: Name or service not known (-2) Kind regards, Roman > -- Original Message -- > From: roman.me...@gismap.ch > To: Wolfgang Schweer , "debian-edu@lists.debian.org" > > Date: 04/21/2021 10:33 AM > Subject: Re: Client time/date synchronisation > > > Hi Wolfgang, > > I just checked the content of /etc/ntp.conf of my "diskless workstation" > client. No line "server ntp" has been appended. > > Guess that explains the problem I'm having, right? > > Kind regards, > Roman > > > On 04/20/2021 5:28 PM Wolfgang Schweer wrote: > > > > > > Hi Roman, > > > > [ roman.me...@gismap.ch, 2021-04-20 ] > > > Fact is that my LTSP clients don't have the correct time if the server > > > has no internet connection, which may be perfectly normal, right? > > > > No, it should work. > > > > > There's none on the client in /opt/ltsp/i386/etc/ntp.conf though. > > > >
Fwd: i386 clients not booting as thin client
Hi folks, I'm still stuck here not being able to boot my clients as "thin client". Any ideas what is causing this? Kind regards, Roman > -- Original Message -- > From: roman.me...@gismap.ch > To: "debian-edu@lists.debian.org" > Date: 04/20/2021 9:24 AM > Subject: i386 clients not booting as thin client > > > Hi folks, > > I'm getting my feet wet with the latest version of Debian Edu 10.9.0. > > I was able to prepare the server and additionally create the image for our > i386 clients. > > All our clients are booting nicely using primary and secondary subnet. > However they only boot as "diskless workstation". In older versions of Debian > Edu I was able to edit file /var/lib/tftpboot/ltsp/i386/lts.conf and add > FAT_RAM_THRESHOLD=1000 for all clients having less than 1GB RAM to boot as > thin client instead of diskless workstation. This is not working for me > anymore. > > I also tried to edit LDAP settings using ldapvi. I added the following entry: > ltspConfig: FAT_RAM_THRESHOLD=1000 > This also doesn't work for me. > > Any ideas how to fix this? > > Kind regards, > Roman
Re: Client time/date synchronisation
Hi Wolfgang, I just checked the content of /etc/ntp.conf of my "diskless workstation" client. No line "server ntp" has been appended. Guess that explains the problem I'm having, right? Kind regards, Roman > On 04/20/2021 5:28 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-04-20 ] > > Fact is that my LTSP clients don't have the correct time if the server > > has no internet connection, which may be perfectly normal, right? > > No, it should work. > > > There's none on the client in /opt/ltsp/i386/etc/ntp.conf though. > > LTSP modifies the setting at client startup. A line 'server ntp' should > be appended at the end of ntp.conf. The value 'ntp' is fetched from LDAP > (ltspConfig: TIMESERVER=ntp). > > Replace ntp with auto and check if it makes a difference (less flexible, > but no name resolution needed in this case). > > Wolfgang
Re: Client time/date synchronisation
Hi Wolfgang, I'm not sure I fully understand the content of the file you sent. Fact is that my LTSP clients don't have the correct time if the server has no internet connection, which may be perfectly normal, right? Kind regards, Roman > On 04/20/2021 12:13 PM Wolfgang Schweer wrote: > > > LTSP switched to systemd-timesyncd, configuration is done in LDAP, see: > > https://salsa.debian.org/debian-edu/debian-edu-config/-/blob/buster/cf3/cf.ntp > > Wolfgang > > (sent from mobile device) > > Am April 20, 2021 9:30:47 AM UTC schrieb roman.me...@gismap.ch: > >Hi Petter, > > > >I found a server line on the server, which is: > >server 127.127.1.0 > > > >There's none on the client in /opt/ltsp/i386/etc/ntp.conf though. > > > >Kind regards, > >Roman > > > >> On 04/20/2021 10:02 AM Petter Reinholdtsen wrote: > >> > >> > >> [Roman Meier] > >> > Is there another way to synchronize the time e.g. with the server > >> > instead of the internet? > >> > >> In earlier times, all clients had their /etc/ntp.conf file set up to > >> synchronise their clocks with the server, for just this kind of > >> scenario. Is this no longer the case? > >> > >> Look for the 'server' lines on the client and server. > >> > >> -- > >> Happy hacking > >> Petter Reinholdtsen
Re: Client time/date synchronisation
Hi Petter, I found a server line on the server, which is: server 127.127.1.0 There's none on the client in /opt/ltsp/i386/etc/ntp.conf though. Kind regards, Roman > On 04/20/2021 10:02 AM Petter Reinholdtsen wrote: > > > [Roman Meier] > > Is there another way to synchronize the time e.g. with the server > > instead of the internet? > > In earlier times, all clients had their /etc/ntp.conf file set up to > synchronise their clocks with the server, for just this kind of > scenario. Is this no longer the case? > > Look for the 'server' lines on the client and server. > > -- > Happy hacking > Petter Reinholdtsen
Client time/date synchronisation
Hi folks, Many of our clients have set the wrong time/date in the bios mostly as a result of issues with old cmos batteries. When the server is connected to the internet the time/date is automatically corrected and things work just fine. However, most of the time internet access is off. Then these clients have the wrong time/date and consequently issues with certificate warnings when using the browsers. Is there another way to synchronize the time e.g. with the server instead of the internet? Kind regards, Roman
i386 clients not booting as thin client
Hi folks, I'm getting my feet wet with the latest version of Debian Edu 10.9.0. I was able to prepare the server and additionally create the image for our i386 clients. All our clients are booting nicely using primary and secondary subnet. However they only boot as "diskless workstation". In older versions of Debian Edu I was able to edit file /var/lib/tftpboot/ltsp/i386/lts.conf and add FAT_RAM_THRESHOLD=1000 for all clients having less than 1GB RAM to boot as thin client instead of diskless workstation. This is not working for me anymore. I also tried to edit LDAP settings using ldapvi. I added the following entry: ltspConfig: FAT_RAM_THRESHOLD=1000 This also doesn't work for me. Any ideas how to fix this? Kind regards, Roman
Epoptes broadcasting using thin client
Hi folks, I'm currently using thin client only. Assisting the students using Epoptes works nicely. However, anytime I try to use either the fullscreen or windowed broadcast, nothing shows on any of the selected clients. Was anyone able to make Epoptes broadcasting work on Debian Edu? Any hints why this is happening and how I can fix it? Kind regards, Roman
Fwd: Re: Boot issue on older hardware using buster
Hi folks, I can come up with new findings. The log is showing weird messages. I can't say why they were not showing in the log submitted earlier. Please have a look! [...] Apr 4 09:48:18 tjener dhcpd[2022]: DHCPDISCOVER from 00:80:45:13:6e:cb via eth0 Apr 4 09:48:19 tjener dhcpd[2022]: DHCPOFFER on 10.0.16.41 to 00:80:45:13:6e:cb via eth0 Apr 4 09:48:20 tjener dhcpd[2022]: DHCPREQUEST for 10.0.16.41 (10.0.2.2) from 00:80:45:13:6e:cb via eth0 Apr 4 09:48:20 tjener dhcpd[2022]: DHCPACK on 10.0.16.41 to 00:80:45:13:6e:cb via eth0 Apr 4 09:48:21 tjener in.tftpd[9950]: tftp: client does not accept options Apr 4 09:48:21 tjener in.tftpd[9956]: tftp: client does not accept options Apr 4 09:48:44 tjener dhcpd[2022]: reuse_lease: lease age 24 (secs) under 25% threshold, reply with unaltered, existing lease for 10.0.16.41 Apr 4 09:48:44 tjener dhcpd[2022]: DHCPDISCOVER from 00:80:45:13:6e:cb via eth0 Apr 4 09:48:44 tjener dhcpd[2022]: DHCPOFFER on 10.0.16.41 to 00:80:45:13:6e:cb via eth0 Apr 4 09:48:44 tjener dhcpd[2022]: reuse_lease: lease age 24 (secs) under 25% threshold, reply with unaltered, existing lease for 10.0.16.41 Apr 4 09:48:44 tjener dhcpd[2022]: DHCPREQUEST for 10.0.16.41 (10.0.2.2) from 00:80:45:13:6e:cb via eth0 Apr 4 09:48:44 tjener dhcpd[2022]: DHCPACK on 10.0.16.41 to 00:80:45:13:6e:cb via eth0 Apr 4 09:48:44 tjener nbd_server[1522]: Spawned a child process Apr 4 09:48:44 tjener nbd_server[10035]: virtstyle ipliteral Apr 4 09:48:44 tjener nbd_server[10035]: connect from 10.0.16.41, assigned file is /opt/ltsp/images/i386.img Apr 4 09:48:44 tjener nbd_server[10035]: Can't open authorization file /etc/ltsp/nbd-server.allow (No such file or directory). Apr 4 09:48:44 tjener nbd_server[10035]: Size of exported file/device is 3731148800 Apr 4 09:48:44 tjener nbd_server[10035]: Starting to serve Apr 4 09:48:51 tjener nbd_server[1522]: Spawned a child process Apr 4 09:48:51 tjener nbd_server[10060]: virtstyle ipliteral Apr 4 09:48:51 tjener nbd_server[10060]: connect from 10.0.16.41, assigned file is /tmp/nbd-swap/10.0.16.41 Apr 4 09:48:51 tjener nbd_server[10060]: Can't open authorization file /etc/ltsp/nbd-server.allow (No such file or directory). Apr 4 09:48:51 tjener nbd_server[10060]: Size of exported file/device is 536870912 Apr 4 09:48:51 tjener nbd_server[10060]: Starting to serve Feb 14 10:12:05 am-008045136ecb systemd-modules-load[556]: Inserted module 'lp' Feb 14 10:12:05 am-008045136ecb systemd-sysusers[567]: Creating group systemd-coredump with gid 999. Feb 14 10:12:05 am-008045136ecb systemd-sysusers[567]: Creating user systemd-coredump (systemd Core Dumper) with uid 999 and gid 999. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:1] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher → /run/speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:2] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.cache → /run/speech-dispatcher/.cache; please update the tmpfiles.d/ drop-in file accordingly. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:3] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.speech-dispatcher → /run/speech-dispatcher/.speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:4] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.cache/speech-dispatcher → /run/speech-dispatcher/.cache/speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:5] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/log → /run/speech-dispatcher/log; please update the tmpfiles.d/ drop-in file accordingly. Feb 14 10:12:05 am-008045136ecb systemd[1]: Starting Flush Journal to Persistent Storage... Feb 14 10:12:05 am-008045136ecb systemd[1]: Started Flush Journal to Persistent Storage. Feb 14 10:12:05 am-008045136ecb systemd[1]: Starting Create Volatile Files and Directories... Feb 14 10:12:05 am-008045136ecb systemd-udevd[586]: Using default interface naming scheme 'v240'. Feb 14 10:12:05 am-008045136ecb systemd-udevd[586]: link_config: autonegotiation is unset or enabled, the speed and duplex are not writable. Feb 14 10:12:05 am-008045136ecb systemd-tmpfiles[599]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:1] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher → /run/speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Feb
Re: Boot issue on older hardware using buster
Hi Wolfgang, I'll have another look at my setup. I will report back if I find something. Thanks a lot for your support! Kind regards, Roman > On 03/19/2021 1:58 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-19 ] > > > The test server I'm using has seen quite some tweaking here and there. > > I'm not very sure if I didn't break something causing this issue. > > That might explain it. I've setup a virtual Buster test environment and > did the changes required to be able to choose a 32-bit thin client from > the PXE menu on the primary subnet. Everything works like expected. > > BTW, I'v updated the instructions on the wiki, see: > https://wiki.debian.org/DebianEdu/Documentation/Buster/HowTo/NetworkClients#Add_LTSP_chroot_to_support_32-bit-PC_clients > > > Reason for the missing NBD image has been a to-and-fro change NFS<->NBD > during development leaving behind a missing comment character in > /etc/ltsp/ltsp-build-cilent.conf. On the command line, the additional > parameter --squash-image overrides the wrong setting > SQUASHFS_IMAGE="False" in ltsp-build-cilent.conf. > > > Anyway, please have a look at the following log. Personally I can't > > find something alarming. My experience is very limited though. > > Thanks for the log file. I've taken a look but can't find anything that > would explain the failure. Maybe someone else would find something. > > Wolfgang
Re: Boot issue on older hardware using buster
Hi Wolfgang, The test server I'm using has seen quite some tweaking here and there. I'm not very sure if I didn't break something causing this issue. Anyway, please have a look at the following log. Personally I can't find something alarming. My experience is very limited though. The log shows all messages when I do the following steps: 1) Start the client 2) Select "thin client" in the pxe menu 3) Login using username "kwepan" 4) The spinner appears and after about 30s I'm redirected back to the login page. Kind regards, Roman [...] Mar 19 06:50:51 tjener dhcpd[1882]: DHCPOFFER on 10.0.16.159 to 00:80:45:13:6e:cb via eth0 Mar 19 06:50:52 tjener dhcpd[1882]: DHCPREQUEST for 10.0.16.159 (10.0.2.2) from 00:80:45:13:6e:cb via eth0 Mar 19 06:50:52 tjener dhcpd[1882]: Wrote 138 leases to leases file. Mar 19 06:50:52 tjener dhcpd[1882]: DHCPACK on 10.0.16.159 to 00:80:45:13:6e:cb via eth0 Mar 19 06:50:52 tjener in.tftpd[22455]: tftp: client does not accept options Mar 19 06:50:52 tjener in.tftpd[22461]: tftp: client does not accept options Mar 19 06:51:22 tjener dhcpd[1882]: reuse_lease: lease age 30 (secs) under 25% threshold, reply with unaltered, existing lease for 10.0.16.159 Mar 19 06:51:22 tjener dhcpd[1882]: DHCPDISCOVER from 00:80:45:13:6e:cb via eth0 Mar 19 06:51:22 tjener dhcpd[1882]: DHCPOFFER on 10.0.16.159 to 00:80:45:13:6e:cb via eth0 Mar 19 06:51:22 tjener dhcpd[1882]: reuse_lease: lease age 30 (secs) under 25% threshold, reply with unaltered, existing lease for 10.0.16.159 Mar 19 06:51:22 tjener dhcpd[1882]: DHCPREQUEST for 10.0.16.159 (10.0.2.2) from 00:80:45:13:6e:cb via eth0 Mar 19 06:51:22 tjener dhcpd[1882]: DHCPACK on 10.0.16.159 to 00:80:45:13:6e:cb via eth0 Mar 19 06:51:22 tjener nbd_server[1361]: Spawned a child process Mar 19 06:51:22 tjener nbd_server[22551]: virtstyle ipliteral Mar 19 06:51:22 tjener nbd_server[22551]: connect from 10.0.16.159, assigned file is /opt/ltsp/images/i386.img Mar 19 06:51:22 tjener nbd_server[22551]: Can't open authorization file /etc/ltsp/nbd-server.allow (No such file or directory). Mar 19 06:51:22 tjener nbd_server[22551]: Size of exported file/device is 3731148800 Mar 19 06:51:22 tjener nbd_server[22551]: Starting to serve Mar 19 06:51:28 tjener nbd_server[1361]: Spawned a child process Mar 19 06:51:28 tjener nbd_server[22575]: virtstyle ipliteral Mar 19 06:51:28 tjener nbd_server[22575]: connect from 10.0.16.159, assigned file is /tmp/nbd-swap/10.0.16.159 Mar 19 06:51:28 tjener nbd_server[22575]: Can't open authorization file /etc/ltsp/nbd-server.allow (No such file or directory). Mar 19 06:51:29 tjener nbd_server[22575]: Size of exported file/device is 536870912 Mar 19 06:51:29 tjener nbd_server[22575]: Starting to serve Mar 19 06:51:36 am-008045136ecb systemd-modules-load[553]: Inserted module 'lp' Mar 19 06:51:36 am-008045136ecb systemd-sysusers[569]: Creating group systemd-coredump with gid 999. Mar 19 06:51:36 am-008045136ecb systemd-sysusers[569]: Creating user systemd-coredump (systemd Core Dumper) with uid 999 and gid 999. Mar 19 06:51:36 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:1] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher → /run/speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Mar 19 06:51:36 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:2] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.cache → /run/speech-dispatcher/.cache; please update the tmpfiles.d/ drop-in file accordingly. Mar 19 06:51:36 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:3] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.speech-dispatcher → /run/speech-dispatcher/.speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Mar 19 06:51:36 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:4] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/.cache/speech-dispatcher → /run/speech-dispatcher/.cache/speech-dispatcher; please update the tmpfiles.d/ drop-in file accordingly. Mar 19 06:51:36 am-008045136ecb systemd-tmpfiles[574]: [/usr/lib/tmpfiles.d/speech-dispatcher.conf:5] Line references path below legacy directory /var/run/, updating /var/run/speech-dispatcher/log → /run/speech-dispatcher/log; please update the tmpfiles.d/ drop-in file accordingly. Mar 19 06:51:36 am-008045136ecb systemd[1]: Starting Flush Journal to Persistent Storage... Mar 19 06:51:36 am-008045136ecb systemd[1]: Started Flush Journal to Persistent Storage. Mar 19 06:51:36 am-008045136ecb systemd[1]: Starting Create Volatile Files and Directories... Mar 19 06:51:36 am-008045136ecb systemd-udevd[590]: Using default interface naming scheme 'v240'. Mar 19 06:51:36 am-008045136ecb
Re: Boot issue on older hardware using buster
Hi Wolfgang, Nope, I tried various users and none worked on "thin client" but worked just fine on "diskless workstation". What else could it be? Kind regards, Roman > On 03/17/2021 11:17 AM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-17 ] > > Another question: > > > > I'm not sure if it's related to the above though. > > > > To login using "diskless workstation" works just fine but I can't > > using "thin client" and I'm redirected back to the login screen. > > This issue is unrelated to the reported one. Most probably you are > trying to login as the same user already logged in on the LTSP server. > Log out on the server and you should be able to log in via the thin > client. > > Wolfgang
Re: Boot issue on older hardware using buster
Hi Roland, Yep, you are right! It seems you had the same problem except for my thin client login issue, that I'm still having. Kind regards, Roman > On 03/16/2021 4:14 PM Roland F.Teichert wrote: > > > Hi Roman, > > I had nearly the same problem. Please read the mails with the subject: > > 'Problem: Debian-Edu i386 ThinClients' > > It's really tricky, but at the end it works well. > > Greetings, > Roland > > Am 16.03.21 um 15:25 schrieb roman.me...@gismap.ch: > > Hi Wolfgang, > > > > I did the changes as suggested, however the same message is thrown. > > > > More ideas? ;) > > > > Kind regards, > > Roman > > > >> On 03/16/2021 11:51 AM Wolfgang Schweer wrote: > >> > >> > >> Hi Roman, > >> > >> [ roman.me...@gismap.ch, 2021-03-16 ] > >>> Negotiation: ..Error: Unknown error returned by server. Server said: > >>> Export unknown Exiting. > >>> > >>> Could it be that the script is somehow broken? > >> > >> Most probably the reported error isn't script related. > >> > >> Add a line to /etc/exports for the 1386 chroot (like the existing line > >> for amd64). > >> > >> Then run 'exportfs -rav' for the change to take effect. > >> > >> Hope it helps, > >> Wolfgang > >
Re: Boot issue on older hardware using buster
Hi Wolfgang, Great! This was it! All clients are booting now! Another question: I'm not sure if it's related to the above though. To login using "diskless workstation" works just fine but I can't using "thin client" and I'm redirected back to the login screen. How can I fix this? Kind regards, Roman > On 03/16/2021 4:15 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-16 ] > > I did the changes as suggested, however the same message is thrown. > > > > More ideas? ;) > > Yes: run 'ltsp-update-image'. > > Wolfgang
Re: Boot issue on older hardware using buster
Hi Wolfgang, I did the changes as suggested, however the same message is thrown. More ideas? ;) Kind regards, Roman > On 03/16/2021 11:51 AM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-16 ] > > Negotiation: ..Error: Unknown error returned by server. Server said: > > Export unknown Exiting. > > > > Could it be that the script is somehow broken? > > Most probably the reported error isn't script related. > > Add a line to /etc/exports for the 1386 chroot (like the existing line > for amd64). > > Then run 'exportfs -rav' for the change to take effect. > > Hope it helps, > Wolfgang
Re: Boot issue on older hardware using buster
Hi Wolfgang, I uncommented the line and rebuilt the pxe menu. Now, when I boot the client I get the following error: Negotiation: ..Error: Unknown error returned by server. Server said: Export unknown Exiting. Could it be that the script is somehow broken? Kind regards, Roman > On 03/15/2021 4:06 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-15 ] > > > I was able to create the new image for i386 devices. I also did the > > changes using ldapvi but it seems to me that only devices connected to > > secondary subnets are handled here. > > Yes. These are the dedicated LTSP networks. > > > Currently I'm interested in switching to the i386 kernel for the > > primary subnet only. > > > > I therefore adjusted the following two files: > > > > /var/lib/tftpboot/debian-edu/default-diskless.cfg > > /var/lib/tftpboot/debian-edu/default-thin.cfg > > If I remember correctly, you need to uncomment the #ltsparch=i386 entry > in /etc/debian-edu/pxeinstall.conf and then run 'debian-edu-pxeinstall' > to regenerate the files below /var/lib/tftpboot/debian-edu/ for the > change from amd64 to i386 for LTSP clients to take effect on the > backbone subnet. > > Wolfgang
Fwd: Re: Boot issue on older hardware using buster
Hi folks, I was able to create the new image for i386 devices. I also did the changes using ldapvi but it seems to me that only devices connected to secondary subnets are handled here. Currently I'm interested in switching to the i386 kernel for the primary subnet only. I therefore adjusted the following two files: /var/lib/tftpboot/debian-edu/default-diskless.cfg /var/lib/tftpboot/debian-edu/default-thin.cfg However, I still get the old message upon booting with an older device. What am I missing here? Kind regards, Roman > -- Original Message -- > From: roman.me...@gismap.ch > To: debian-edu@lists.debian.org > Date: 03/14/2021 4:11 PM > Subject: Re: Boot issue on older hardware using buster > > > Hi Wolfgang, > > Thanks for the fast reply! > > That's exactly right! Both, "diskless workstation" and "thin client" can't > boot anymore. > > Documentation seems rather straight forward. I'll give a try. ;) > > Kind regards, > Roman > > > On 03/14/2021 1:04 PM Wolfgang Schweer wrote: > > > > > > Hi Roman, > > > > [ roman.me...@gismap.ch, 2021-03-14 ] > > > > > We recently updated our Debian Edu server from Stretch to Buster. We > > > use a number of older Panasonic Toughbooks and all of them are no > > > longer able to boot. > > > > > > The following message is thrown on those Toughbooks: > > > > > > This kernel requires an x86-64 CPU, but only detected an i686 CPU. > > > Unable to boot - please use a kernel appropriate for your CPU. > > > > I assume the Toughbooks are used as LTSP clients and the Buster main > > server has been a new installation, not an upgrade from Stretch. See the > > manual how to support 32-bit clients: > > https://wiki.debian.org/DebianEdu/Documentation/Buster/HowTo/NetworkClients#Add_LTSP_chroot_to_support_32-bit-PC_clients > > > > Wolfgang
Re: Boot issue on older hardware using buster
Hi Wolfgang, Thanks for the fast reply! That's exactly right! Both, "diskless workstation" and "thin client" can't boot anymore. Documentation seems rather straight forward. I'll give a try. ;) Kind regards, Roman > On 03/14/2021 1:04 PM Wolfgang Schweer wrote: > > > Hi Roman, > > [ roman.me...@gismap.ch, 2021-03-14 ] > > > We recently updated our Debian Edu server from Stretch to Buster. We > > use a number of older Panasonic Toughbooks and all of them are no > > longer able to boot. > > > > The following message is thrown on those Toughbooks: > > > > This kernel requires an x86-64 CPU, but only detected an i686 CPU. > > Unable to boot - please use a kernel appropriate for your CPU. > > I assume the Toughbooks are used as LTSP clients and the Buster main > server has been a new installation, not an upgrade from Stretch. See the > manual how to support 32-bit clients: > https://wiki.debian.org/DebianEdu/Documentation/Buster/HowTo/NetworkClients#Add_LTSP_chroot_to_support_32-bit-PC_clients > > Wolfgang
Boot issue on older hardware using buster
Hi folks, We recently updated our Debian Edu server from Stretch to Buster. We use a number of older Panasonic Toughbooks and all of them are no longer able to boot. The following message is thrown on those Toughbooks: This kernel requires an x86-64 CPU, but only detected an i686 CPU. Unable to boot - please use a kernel appropriate for your CPU. Can this be fixed? What's the best approach for this? Kind regards, Roman
Re: Server not booting anymore
> Generic (Debian) advice (works for a lot of other Linux distros too) > The file /etc/fstab contains the filesystems to be mounted upon boot. [...] > UUID="0e799f6c-a675-48da-bb46-f51128a8e3a7" TYPE="swap" > > (note that PARTUUID is a different thing, you almost always want UUID) > Armed with this information, you should be able to figure out what has > happened: > - the partition in question got a new UUID (this is often a bad sign, > the UUID is created when the partition is created) > - the /etc/fstab got changed somehow; simply change the correct line > to point to the correct UUID > > Apologies if this is well known already. > If not, hope it helps. > -- > Regards, > Torfinn Ingolfsen > > Hi folks, Thanks so much for your contributions! I was finally able to fix the issue! It seems that the liveCd I used to boot Debian-Edu did some changes to fstab. I'm not sure tho. Fact is that I found some funny content in my fstab: #Entry for /dev/loop0 : UUID=2020-02-08-14-49-26-00 /media/roman/d-live\040nf\04010.3.0\040xf\040amd64 iso9660 ro,nosuid,nodev,relatime,nojoliet,check=s,map=n,blocksize=2048, uid=1000,gid=1000, dmode=500,fmode=400 0 0 I also found the following file: /etc/fstab/fstab-disk-manager-save The file date is equal to the time this boot issue started. I suspect that it's some kind of backup file. The following step fixed the issue: cp fstab-disk-manager-save fstab Thanks again!!! Kind regards, Roman
Re: Server not booting anymore
> On Thu, May 07, 2020 at 03:30:08PM +0200, Roman Meier wrote: >> I rather seriously messed up my Debian Edu server this morning! >> >> I booted the server using a Debian-liveCd and it seems this did some >> harm >> to grub for reasons unknown since I didn't do any installation or >> changes >> to the system I'm aware of. > [..] >> Any ideas how to fix this? > > You could try to use the Debian Installer's rescue mode, see: > https://wiki.debian.org/DebianEdu/Documentation/Buster/Installation#Installation_types_and_options > (Advanced options -> Rescue mode) to inspect the broken system, i.e: > just use the Debian Edu Buster Netinst ISO to boot your main server, > see: https://wiki.debian.org/DebianEdu/Status/Buster#Download_using_http > > Good luck, > Wolfgang > That's working nicely. Don't have a clue on how to fix it tho. :(
Server not booting anymore
Hi folks, I rather seriously messed up my Debian Edu server this morning! I booted the server using a Debian-liveCd and it seems this did some harm to grub for reasons unknown since I didn't do any installation or changes to the system I'm aware of. Anyway the following shows now when I'm booting: [...] /dev/mapper/vg_system-root: clean, 111059/921344 files, 2855990/3050240 blocks Gave up waiting for /usr file system device. Common problems: - Boot args (cat /proc/cmdline) - Check rootdelay= (did the system wait long enough?) - Missing modules (cat /proc/modules: ls /dev) ALERT! UUID=378.b07 does not exist. Dropping to shell! BusyBox v1.30.1 (Debian 1:1.30.1-4) buil-in shell (ash) Enter 'help' for a list of built-in commands. (initramfs) _ Any ideas how to fix this? Kind regards, Roman
Re: Backup
Hi Frans, Yep, you were right. The client setup was entirely missing and therefore no restore was possible using the web interface. I was able to draw inspiration from the example configuration file you mentioned and our backup is working now. However, the configuration file seems not really adequate for production use of a Debian Edu server. Can you either share a better configuration file or tell me, what folders are important to backup to be save and have all important files? I'm more or less using a standard configuration of Debian Edu. Kind regards, Roman > Hi Roman, > > Roman Meier schreef op do 20-02-2020 om 08:25 [+0100]: >> Hi folks, >> >> After having lost some data I have started to read more about the backup >> feature included in Debian Edu. >> >> The backups on our server seem to run just fine according to the status >> page of slbackup.php. However, menuitems "Restore" and "Maintenance" >> give >> no options to choose from. Why? > > Most likely your /etc/slbackup/slbackup.conf configuration file lay-out is > the reason why. > >> >> I also wonder about the "Config" page. Fields under "Server config" are >> filled but none are set in "Client config". I don't actually understand >> the difference between the 2 sections. I couldn't find any >> documentation. > > The package comes with an example configuration file. Take a look at > /usr/share/doc/slbackup/examples/slbackup-server+ltsp.conf > and compare it to your own /etc/slbackup/slbackup.conf configuration file. > > The slbackup system is an ordinary client server system. > The main server tjener runs the back up service and thus acts as the > slbackup server. > Machines that have some files and directories backed up are clients. Since > also files and directories of tjener are backed up by the slbackup > service, > tjener is in this respect the backup client. If you have other machines in > your network that are backed up, they also are clients. > >> >> Kind regards, >> Roman >> > > -- > Kind regards, > Frans Spiesschaert > > >
Re: How to restore the newteacher template in Gosa2?
Hi Wolfgang, I was finally able to restore the teacher template with your help. It seems to work fine. Thanks! Kind regards, Roman P.S.: This taught me a lesson to always ensure having a working backup. :) > Hi Roman, > > On Thu, Feb 20, 2020 at 09:03:29AM +0100, Roman Meier wrote: >> Seems this is a little tricky. >> >> It appears I actually deleted user "NewTeacher". Did this result in the >> deletion of its group, which may have been "newteacher"? Do I need to >> recreate this group as well? > > Yes. The 'newteacher' group is similar to the 'newstudent' group. GID > for newteacher should be 1002 as opposed to 1001 for newstudent. > > Good luck, > Wolfgang >
Re: How to restore the newteacher template in Gosa2?
Hi Wolfgang, Seems this is a little tricky. It appears I actually deleted user "NewTeacher". Did this result in the deletion of its group, which may have been "newteacher"? Do I need to recreate this group as well? Wish I had a backup. Such a mess :( Kind regards, Roman > On Tue, Feb 18, 2020 at 08:43:40AM +0100, Roman Meier wrote: >> Unfortunately I managed to accidentally delete the newteacher template >> in >> Gosa2. >> >> How can I restore it? > > You should e able to re-create it. Take a look at the NewStudent > template (main tab, POSIX and Samba tab). The NewTeacher template is > similar, just replace student(s) with teacher(s) in all fields. IIRC, > the only difference is the UID/GID which is 1001 as opposed to 1002 for > the students and the group membership (teachers und students). Just > compare an existing teacher account to be sure about it. > > Good luck, > Wolfgang >
Re: How to restore the newteacher template in Gosa2?
Hi Frans, That's great news! However, I come to realize that our server was not configured correctly for our use case. Backups are configured by default to run at 1am in the morning. In our school we always off the server over night and consequently no backup was ever made. :( I'm now in the process to read more about backupping to make sure this will proparly in the future. Kind regards, Roman > Wolfgang Schweer schreef op di 18-02-2020 om 21:59 [+0100]: >> On Tue, Feb 18, 2020 at 08:43:40AM +0100, Roman Meier wrote: >> > Unfortunately I managed to accidentally delete the newteacher template >> > in >> > Gosa2. >> > >> > How can I restore it? >> >> You should e able to re-create it. Take a look at the NewStudent >> template (main tab, POSIX and Samba tab). The NewTeacher template is >> similar, just replace student(s) with teacher(s) in all fields. IIRC, >> the only difference is the UID/GID which is 1001 as opposed to 1002 for >> the students and the group membership (teachers und students). Just >> compare an existing teacher account to be sure about it. >> > The main server runs a backup service. See section "11.2. Backup > Management" of the manual on that topic. So you could restore a backup > copy > of the file /etc/ldap/gosa.ldif from the day before you made your mistake. > That is the file that contains the newteacher template. > > -- > Kind regards, > Frans Spiesschaert > > >
Backup
Hi folks, After having lost some data I have started to read more about the backup feature included in Debian Edu. The backups on our server seem to run just fine according to the status page of slbackup.php. However, menuitems "Restore" and "Maintenance" give no options to choose from. Why? I also wonder about the "Config" page. Fields under "Server config" are filled but none are set in "Client config". I don't actually understand the difference between the 2 sections. I couldn't find any documentation. Kind regards, Roman
How to restore the newteacher template in Gosa2?
Hi folks, Yesterday I had to register a new teacher in our school. Unfortunately I managed to accidentally delete the newteacher template in Gosa2. How can I restore it? Kind regards, Roman
Re: Debian Edu and Raspberry Pi
Hi Roland, Thanks! Interesting approach and documentation! :) Seems there's support for pxe boot without the need for sd cards starting from rpi 3. Could this work with Debian Edu as well? How to do power your rpi's? Each having an individual power supply? Do you have any experiences with PoE/shared power supplies? Kind regards, Roman > Hi Roman, > > I' m using RPI 2 with 1 GB Ram as ThinClient in my DebianEdu Network > and the RPI works pretty fine. All you need is a special Raspbian an > the sd card, which managed the pxe boot, and everything works fine. > > http://rpitc.blogspot.com/ > > Greetings, > Roland > > On 08.02.20 13:24, Roman Meier wrote: >> Hi folks, >> >> I'm looking for hardware alternatives to be used as clients. Currently >> we >> are using Toughbooks (CP28 and 29) as thin clients. >> >> I am considering the Raspberry Pi but information is very limited. >> However, I came across the following documentation, which sounds >> promising: >> >> https://wiki.debian.org/DebianEdu/HowTo/RaspberryPiWorkstation >> >> How good or bad are experiences with Debian Edu running on Raspberry >> Pis? >> Is this a viable option? >> >> The documentation from above installs a full workstation requiring an Sd >> card. Is the Rpi strong enough to handle this well, e.g. using big >> softwares like libreOffice? >> >> Is it possible to use the Rpi as thin client instead? >> >> Kind regards, >> Roman >> > >
Re: Roaming workstation installation
Hi Wolfgang That was exactly the problem here. I fixed it issuing the following two commands on the server: apt update apt upgrade The server reboot was a little scary with various never seen notices and an unexpected halt. Anyway, I rebooted again and it booted nicely. After this the roaming workstation installation went through nicely with no issues. I was a little surprised tho that it had to download an extra 3gb of data. I was expecting that most data had been downloaded already and cached on the server. Is that normal? Anyway, it works nicely so far. :) Kind regards, Roman > Moin Roman, > > On Sun, Feb 16, 2020 at 03:51:52PM +0100, Roman Meier wrote: >> Download installer components >> >> No kernel modules were found. This probably is due to a mismatch between >> the kernel used by this version of the installer and the kernel version >> available in the archive. > > Point releases for both Stretch and Buster happened last week > (2010-02-08) - including netboot related changes. > > Maybe you missed to upgrade the main server? > > Wolfgang >
Roaming workstation installation
Hi folks, I tried to do my first installation of a roaming workstation this weekend. I therefore connected the laptop to the school network and the pxe menu showed up as expected. I chose the 4th entry which was "Install Edu/amd64 (xfce)". Things moved on smoothly until I got the following notice: ** Download installer components No kernel modules were found. This probably is due to a mismatch between the kernel used by this version of the installer and the kernel version available in the archive. If you're installing from a mirror, you can work around this problem by choosing to install a different version of Debian. The install will probably fail to work if you continue without kernel modules. Continue to install without loading kernel modules? ** I tried various things to fix this but always ultimately failed. Any ideas what's the best approach is to get past this issue? Kind regards, Roman
Re: Debian Edu and Raspberry Pi
Using Rpi's and LTSP isn't something new I think. At least that's the impression I got glancing through https://pi-ltsp.net/ and http://pinet.org.uk/. :) >> A thin client solution may >>eliminate the costs for sd cards and power supplies and maybe there can >>even be a way to switch to a Rpi Zero instead. > > Rumour has it the new LTSP can boot RPi. > > I will test at Teckids when recreating our LTSP environment. > > -nik > -- > Sendt fra min Android-enhet med K-9 e-post. Unnskyld min kortfattethet. > >
Re: Debian Edu and Raspberry Pi
I'm also looking at this from a cost perspective. A thin client solution based on Rpi's is rather affordable but may still be on the upper side for budgets in schools of third world countries. A thin client solution may eliminate the costs for sd cards and power supplies and maybe there can even be a way to switch to a Rpi Zero instead. > [Roman Meier] >> How good or bad are experiences with Debian Edu running on Raspberry >> Pis? >> Is this a viable option? > > I know the RPi 1 was a little short on memory, it could run as a thin > client and a fairly limited desktop. 512 MiB of RAM is not a lot these > days. I would imagine the RPi 3 would work just fine. I do not believe > anyone tested this the last few years, but the recipe you found should > get you started at least. Perhaps the recipe still work? It try to > replicate the installation procedure run from the normal installer, so > it just might. :) > > -- > Happy hacking > Petter Reinholdtsen > >
Debian Edu and Raspberry Pi
Hi folks, I'm looking for hardware alternatives to be used as clients. Currently we are using Toughbooks (CP28 and 29) as thin clients. I am considering the Raspberry Pi but information is very limited. However, I came across the following documentation, which sounds promising: https://wiki.debian.org/DebianEdu/HowTo/RaspberryPiWorkstation How good or bad are experiences with Debian Edu running on Raspberry Pis? Is this a viable option? The documentation from above installs a full workstation requiring an Sd card. Is the Rpi strong enough to handle this well, e.g. using big softwares like libreOffice? Is it possible to use the Rpi as thin client instead? Kind regards, Roman
Re: Partition resizing fails
Yep, this is exactly what happened. /etc/lvm/archive/.lvm_tjener.intern_6654_928137001: fclose failed: No space left on device :) > Hello, > > > Am Fri, 27 Dec 2019 16:59:49 +0100 > schrieb "Roman Meier" : > >> The issue was that a minimal amount of space is required to execute an >> lvextent of /dev/vg_system/root. In my case the space was just not big >> enough. > > I guess, the LVM tools tried to create a new meta backup below > /etc/lvm/archive > and failed (in a mysterious way). I am not sure, how to work around this > (or > provide a better error message) in this case. > > Lars > >
Re: Partition resizing fails
Hi Wolfgang, I was finally able to fix it. Took me way too long to figure this one out. The issue was that a minimal amount of space is required to execute an lvextent of /dev/vg_system/root. In my case the space was just not big enough. I therefore temporarily moved some files out of /etc/debian-edu/www/. Then I rebooted and the two commands (lvextend and resize2fs) executed nicely. Thanks for your hints and assistance! :) > On Fri, Dec 27, 2019 at 10:25:14AM +0100, Roman Meier wrote: >> That was the first thing I did. :) >> >> Didn't help me tho. :( > >> >> vgdisplay show this: >> >> Free PE / Size 204547 / 799.01 GiB >> >> > /dev/mapper/vg_system-root 2108840 2092452 0 100% >> / >> >> > /dev/mapper/vg_system-usr 15417224 14836932 0 100% >> >> /usr >> >> > >> >> > I tried to increase the size using lvextend but it gives me a no >> space >> >> > left on device. > > I'm just curious how you used lvextend. > > As an example (like the one in the manual) this should add 500 MiB to /: > lvextend -L+500M /dev/vg_system/root > > Then resize it: > resize2fs /dev/vg_system/root > > Wolfgang >
Re: Partition resizing fails
That was the first thing I did. :) Didn't help me tho. :( > On Fri, Dec 27, 2019 at 04:18:23AM +0100, Roman Meier wrote: >> Sorry, I forgot to add that there should be plenty of space left of the >> device. >> >> vgdisplay show this: >> Free PE / Size 204547 / 799.01 GiB >> > /dev/mapper/vg_system-root 2108840 2092452 0 100% / >> > /dev/mapper/vg_system-usr 15417224 14836932 0 100% >> /usr >> > >> > I tried to increase the size using lvextend but it gives me a no space >> > left on device. >> > Any ideas how to fix this? > > See the 'Logical Volume Management' section in the manual: > > https://wiki.debian.org/DebianEdu/Documentation/Buster/HowTo/Administration#Resizing_Partitions > > Wolfgang >
Re: Partition resizing fails
Sorry, I forgot to add that there should be plenty of space left of the device. vgdisplay show this: Free PE / Size 204547 / 799.01 GiB > Hi folks, > > It may be a bit OT but maybe you are the best to know how to fix this. > > I did a number of installations today on my debian edu server and run out > of disc space. > > /dev/mapper/vg_system-root 2108840 2092452 0 100% / > /dev/mapper/vg_system-usr 15417224 14836932 0 100% /usr > > I tried to increase the size using lvextend but it gives me a no space > left on device. > > The same is happening when I run the following: > debian-edu-fsautoresize -n > > root@tjener:/skole/tjener/home0/roman# /sbin/debian-edu-fsautoresize -n > exec: lvextend -L2374451k /dev/vg_system/root > Rounding size to boundary between physical extents: <2.27 GiB. > /etc/lvm/archive/.lvm_tjener.intern_6654_928137001: fclose failed: No > space left on device > error: unable to resize /dev/mapper/vg_system-root > root@tjener:/skole/tjener/home0/roman# /sbin/debian-edu-fsautoresize > exec: lvextend -L2374451k /dev/vg_system/root (dryrun, add -n to > activate) > exec: resize2fs /dev/mapper/vg_system-root (dryrun, add -n to activate) > root@tjener:/skole/tjener/home0/roman# > > Any ideas how to fix this? > > Kind regards, > Roman > >
Partition resizing fails
Hi folks, It may be a bit OT but maybe you are the best to know how to fix this. I did a number of installations today on my debian edu server and run out of disc space. /dev/mapper/vg_system-root 2108840 2092452 0 100% / /dev/mapper/vg_system-usr 15417224 14836932 0 100% /usr I tried to increase the size using lvextend but it gives me a no space left on device. The same is happening when I run the following: debian-edu-fsautoresize -n root@tjener:/skole/tjener/home0/roman# /sbin/debian-edu-fsautoresize -n exec: lvextend -L2374451k /dev/vg_system/root Rounding size to boundary between physical extents: <2.27 GiB. /etc/lvm/archive/.lvm_tjener.intern_6654_928137001: fclose failed: No space left on device error: unable to resize /dev/mapper/vg_system-root root@tjener:/skole/tjener/home0/roman# /sbin/debian-edu-fsautoresize exec: lvextend -L2374451k /dev/vg_system/root (dryrun, add -n to activate) exec: resize2fs /dev/mapper/vg_system-root (dryrun, add -n to activate) root@tjener:/skole/tjener/home0/roman# Any ideas how to fix this? Kind regards, Roman
Connection issue with Xrdp
Hi folks, I did my first installation of Debian Edu 10 a few days ago and I'm currently testing it. I tried to connect to Xrdp using Windows via primary subnet immediately getting the following error message: An authentication error has occurred (Code: 0x80090302). This used to work just fine on my older Server using Debian Edu 9. Though Xrdp is booting and running nicely I suspect that it needs some basic configuration. Any hints? Kind regards, Roman
Missing gcompris package
Hi folks, One of the most popular packages for our younger kids was clearly gcompris. It seems to have been removed from the default installation. Can you tell me what happened? Is it possible to install it? Kind regards, Roman
Re: Is there any program to design a schedule of lessons?
Hi Andreas, As Petter is saying, have a look at fet. Works great for us also due to excellent support in German as well. :) https://lalescu.ro/liviu/fet/ Kind regards, Roman > On Wed, Oct 02, 2019 at 01:37:04PM +0200, Petter Reinholdtsen wrote: >> > just a quick question: Is there any free program to craft >> > a schedule of lessons in a school? >> >> Not quite sure what you mean, but could 'fet' be in the class of program >> you are looking for? It take information about events, resources and >> participants, and try to generate a schedule that fit everyone. In >> other word, generate the time table. > > May be I have a wrong translation. In German it is "Stundenplan" and > means the weekly repeatet sequence of all lessons of a school for all > classes. > > Kind regards > > Andreas. > > -- > http://fam-tille.de > >
Re: How to get downloadable version of this mailing list?
Hi Nik, That's great To have the past 3 years should be perfect since DebianEdu is changing a lot. I don't know about the file size. Maybe you have to make it available as download. Anyway you can try to email it to me. Kind regards, Roman > Since I got all mailinglists separately in mailboxes myself, I can send > you > archives at least of the last two or three years as mbox, if you want ;)
Re: How to get downloadable version of this mailing list?
Thanks, Nik Thats too bad, but I do understand the reasons... Kind regards, Roman > Hi, > >> My internet connection is rather unstable and slow sometimes. >> >> Is there a way I can download the monthly archive files e.g. as mbox >> file >> to be able to search offline? > > sadly not, since some people refused to have their public e-mail published > in 2011: > https://lists.alioth.debian.org/pipermail/soc-coordination/2011-August/001076.html > > -nik > > -- > Dominik George (1. Vorstandsvorsitzender, pädagogischer Leiter) > Teckids e.V. - Erkunden, Entdecken, Erfinden. > https://www.teckids.org/ > >
How to get downloadable version of this mailing list?
Hi folks, My internet connection is rather unstable and slow sometimes. Is there a way I can download the monthly archive files e.g. as mbox file to be able to search offline? Kind regards, Roman
Re: Demonstration and teaching tool
Anybody? Is my question clear? > Hi folks, > > Being able to teach and support an individual or a group of students of a > class should also involve audio. I was thinking of using headsets. Maybe > even using the webcam as well? > > Is there something out there? > > Kind regards, > Roman > >> On Fri, Nov 17, 2017 at 10:50:36AM +, Dominik George wrote: >>> I could place an evaluation of the different available tools on the >>> agenda for one of the next meetings of the German Skolelinux >>> community. This way, we can find a tool that is the best match and >>> collect ideas what would need to be changed to meet both pedagogic and >>> privacy demands. >> >> Sounds like a plan :) >> >> Wolfgang >> > >
Re: Demonstration and teaching tool
Hi folks, Being able to teach and support an individual or a group of students of a class should also involve audio. I was thinking of using headsets. Maybe even using the webcam as well? Is there something out there? Kind regards, Roman > On Fri, Nov 17, 2017 at 10:50:36AM +, Dominik George wrote: >> I could place an evaluation of the different available tools on the >> agenda for one of the next meetings of the German Skolelinux >> community. This way, we can find a tool that is the best match and >> collect ideas what would need to be changed to meet both pedagogic and >> privacy demands. > > Sounds like a plan :) > > Wolfgang >
Re: Demonstration and teaching tool
> Hi, > >> IÂm not very comfortable with the fact that a teacher has access to all >> thin clients on both subnets. If chroot is epoptes enabled then he has >> access to diskless workstations as well enabling him to monitor all LTSP >> clients in the system without any user knowing about it. >> >> Is this not a security issue? > > yes, it is, among other things, like being disrespectful, contradicting > everything we should teach students about digitalisation and privacy and > last, but not least, probably being illegal (German law, for example, > requires all parents *and* students themselves signing a contract > allowing you to spy on them, which most schools I know simply ignore). > > I figure you really only want the demonstration feature, not the spyware > feature, but as a note for everyone stumbling across this thread because > they want the spyware features: > > If managing a classroom without spying on your students is beyond > your capabilities, please go find something to do that fits in with your > capabilities instead of managing classrooms. Thanks. Hi Nik, Thanks for your input! I agree with you! What alternative to Epoptes do you recommend? Kind regards, Roman
Re: xrdp sessions
> On Mon, Nov 06, 2017 at 11:59:40AM +0100, Roman Meier wrote: >> > Currently xrdp offers a rather big selection of sessions: >> > - Xorg >> > - Xvnc >> > - console >> > - vnc-any >> > - sesman-any >> > - rdp-any >> > - neutrinordp-any >> > >> > Is there a way to remove those, which are not needed > > I figure this to be configurable in /etc/xrdp/xrdp.ini > > Just comment out want you dont want, i.e. each line below the [Xorg] > section and see if it works. I commented out all sections except Xorg. Works perfectly!!! Thanks!!!
Re: Demonstration and teaching tool
> On Mon, Nov 06, 2017 at 12:02:39PM +0100, Roman Meier wrote: >> > Our computer lab is entirely made up of debianEdu thin clients. We >> > currently have 20 thin clients using older Panasonic Toughbooks. >> > >> > Im looking for a way a teacher can demonstrate something on his thin >> > client and all students can observe on their thin clients what the >> teacher >> > is actually doing in realtime. >> > >> > How is this done? > > Packages 'epoptes' on the server, 'epoptes-client' on the client side. > Alternatively 'italc-master' and 'italc-client'. > > Never tested Epoptes, though. Last time I used iTALC (> 10 years ago) > it worked but had some issues. Please note that iTALC is discontinued; > the successor Veyon is a complete rewrite and not yet available in > Debian. Hi Wolfgang, I tried Epoptes. It works rather well and gives the teacher a number of helpful tools. However, I wasnt able to make any of the screen broadcasting tools work. Does it work at your end? Im not very comfortable with the fact that a teacher has access to all thin clients on both subnets. If chroot is epoptes enabled then he has access to diskless workstations as well enabling him to monitor all LTSP clients in the system without any user knowing about it. Is this not a security issue? Kind regards, Roman
Re: Demonstration and teaching tool
Anybody? > Hi folks, > > Our computer lab is entirely made up of debianEdu thin clients. We > currently have 20 thin clients using older Panasonic Toughbooks. > > Im looking for a way a teacher can demonstrate something on his thin > client and all students can observe on their thin clients what the teacher > is actually doing in realtime. > > How is this done? > > Kind regards, > Roman > >
Re: Problem connecting to server using remote desktop connection
Anybody? > Hi folks, > > I regularly face the problem that I cant login into the server using > remote desktop connection. > > So far I could always fix this by rebooting the server. > > What can cause this? Is anybody facing this as well? > > Kind regards, > Roman > >
Re: Get a pxe menu for secondary subnet as well
Anybody? > Hi folks, > > By default, the secondary subnet is a subnet dedicated purposely for LTSP > clients. No pxe menu is shown and depending on the available ram of a > client, it is started as thin client or diskless workstation. > > I wish to have a pxe menu for the secondary subnet as well. The main > reason is that I want to make available alternative images e.g. puppy > linux. > > How is this done? > > Kind regards, > Roman > > >
Re: xrdp sessions
Anybody? > Hi folks, > > Currently xrdp offers a rather big selection of sessions: > - Xorg > - Xvnc > - console > - vnc-any > - sesman-any > - rdp-any > - neutrinordp-any > > I wasnt able to make any of them work except for Xorg, which works > beautifully. > > What is the purpose of the others? > > Are they working? > > Are they recommended to be used? > > Is there a way to remove those, which are not needed or shouldnt be used? > > Kind regards, > Roman > >
Demonstration and teaching tool
Hi folks, Our computer lab is entirely made up of debianEdu thin clients. We currently have 20 thin clients using older Panasonic Toughbooks. Im looking for a way a teacher can demonstrate something on his thin client and all students can observe on their thin clients what the teacher is actually doing in realtime. How is this done? Kind regards, Roman
