NAS4Free 11.0.0.4.4249 released (FreeBSD 11.0-RELEASE-P10)

[zoon01]

The NAS4Free Team is pleased to announce the release of version 11.0.0.4.4249 - Sayyadina

Download build 11.0.0.4.4249
Read the readme_11.0

Major Change!

cifs/smb note:
==============
From Samba 4.5.0 and up the default value for the "ntlm auth" option is changed from "yes" to "no".
This may have impact on very old clients which doesn't support NTLMv2 yet.
Only if you really need NTLMv1 Authentication you can set in Additional Parameters: "ntlm auth = yes".

BUILD 11.0.0.4.4249

Changes after release 11.0.0.4.4195
===============================
- Upgrade to FreeBSD 11.0-RELEASE-P10.
- Update translations.
- Add new framework to LAGG.
- Add healthcheck script for zfs.
- Upgrade syncthing to 0.14.27.
- Upgrade tmux to 2.4.
- Upgrade samba to 4.6.3.
- Upgrade nano to 2.8.2.
- Upgrade virtualbox-ose to 5.1.22.
- Fix IE11 and Edge browser does not operate correctly with novnc.

Note:
Clear your browser's cache to avoid display issues after upgrade.
Make a new backup of your system configuration and store this on safe location.

Enjoy!

Regards,
The NAS4Free Team

[ernie]

Hello

Thanks to the dev team for all new releases.

One question : is it possible to have detail on 'Add healthcheck script for zfs' ? What are exactly the tests ?

Thanks

[ms49434]

Hello

Thanks to the dev team for all new releases.

One question : is it possible to have detail on 'Add healthcheck script for zfs' ? What are exactly the tests ?

Thanks

Code: Select all

nas4free: /etc# ./healthcheck.zfs -h
Name:
  healthcheck.zfs
Description:
  Checks the status of all ZFS pools.
  Checks all ZFS pools for read, write and checksum errors.
  Sends an email report.
Options:
  -h : print this help text and exit.
  -s : send success email as well.
  -m address : use alternative email address(es) instead of the email address(es) defined in the WebGUI
nas4free: /etc#

For each pool /sbin/zpool list -H -o health [pool name] is run to determine the pool status.
/sbin/zpool status [pool name] is run and read, write or checksum errors are extracted.
Best would be to create a root cron job for healthcheck.zfs.

A detailed email is sent when a pool status is not ONLINE or read, write or checksum errors have been reported:

Code: Select all

Subject: [ALERT] nas4free.local: ZFS pool health check detected one or more issues.
Sample Body:
The status of ZFS pool "jupiter" is DEGRADED.
ZFS Pool "jupiter" encountered read, write and checksum errors.

  pool: jupiter
 state: DEGRADED
  scan: resilvered 842M in 0h0m with 0 errors on Fri Apr 21 22:39:23 2017
config:

	NAME        STATE     READ WRITE CKSUM
	jupiter     DEGRADED       1     2     1
	  raidz2-0  DEGRADED       1     2     1
	    da2     MISSING       0     0     0
	    da3     ONLINE       1     0     0
	    da4     ONLINE       0     2     1
	    da5     ONLINE       0     0     0
	    da0     ONLINE       0     0     0
	    da1     ONLINE       0     0     0

A short email is sent when all pools are ONLINE and no read, write or checksum errors have been reported and the -s flag is set:

Code: Select all

Subject: [SUCCESS] nas4free.local: ZFS pool health check completed successfully.

Please let me know if you think more checks should be performed or additional information is required.

[Victek]

Hi all,
Thanks for this new version and the dedication for it.I'm reading the following lines after update to this release, I think it's related to new samba version. Any ideas where is the problem source?. Version .4195 was running fine.

Code: Select all

smbd[28971]: reply_sesssetup_and_X: Rejecting attempt at 'normal' session setup after negotiating spnego.
smbd[28971]: [2017/05/05 23:15:43.660413, 0] ../source3/smbd/sesssetup.c:863(reply_sesssetup_and_X)

Cheers

[ernie]

Hello

And concerning
Add new framework to LAGG

What does it mean ?
Better compatibility with lagg in FreeBSD and nas4free ?

Thanks for your inputs

[lindsay]

Is the option WOL removed in newer versions and is spindown taken away?
Do not see WOL option in network and can not get spindown of hdd`s to work.

New install but i was seeing it on an upgraded version?
Is it something i did wrong?

I do have extended GUI installed though

Edit: opt3 did show WOL,strange that others did not thogh

[MikeMac]

I have installed 11.0.0.4.4249 full on zfs mirror, then reboot from USB flash vith same version, embedded.
I try to import zroot pool - but with strange failure


I then change everything - embedded and Live USB version, installed fresh full root on zfs on the different SSD, even try different motherboard - but problem is the same

Code: Select all

nas4free: ~# camcontrol devlist
<Patriot Warp V2 32GB SSD 02.10104>  at scbus0 target 0 lun 0 (pass1,ada0)
<Generic Flash Disk 8.07>          at scbus6 target 0 lun 0 (pass0,da0)
nas4free: ~# zpool status
no pools available
nas4free: ~# zpool import
   pool: zroot
     id: 12484813833227720915
  state: ONLINE
 status: The pool was last accessed by another system.
 action: The pool can be imported using its name or numeric identifier and
        the '-f' flag.
   see: http://illumos.org/msg/ZFS-8000-EY
 config:

        zroot       ONLINE
          ada0p3    ONLINE
nas4free: ~# zpool import -f zroot
nas4free: ~# zpool status
internal error: failed to initialize ZFS library
nas4free: ~# uname -a
FreeBSD nas4free.local 11.0-RELEASE-p10 FreeBSD 11.0-RELEASE-p10 #0 r317803M: Thu May  4 21:48:37 CEST 2017     root@dev.nas4free.org:/usr/obj/nas4free/usr/src/sys/NAS4FREE-amd64  amd64

Also after "internal error: failed to initialize ZFS library" msg
nas4free web gui no more acessible.

UPDATE
I have virtualbox installation for build 11.0.0.4.3330 with rootonzfs and data Pool
I have started this virtual machine from NAS4Free-x64-LiveCD-11.0.0.4.4249 -
Data pool was imported without problem. But then I attempt to import zroot pool - I see
internal error: failed to initialize ZFS library
message

Then I had downloaded FreeBSD-11.0-RELEASE-amd64-disc1.iso and start from it, multiuser. Import zroot, zpool status - internal error: failed to initialize ZFS library
Attempt with singleuser -the same.

Is there something terribly strage with nas4free rootonzfs pool?

[carloskar]

Hi
The sshd_config file is created incorrectly, the ChallengeResponseAuthentication and AllowTcpForwarding is added twice, where the second one is always no.

When this is the content in the file /conf/config.xml:

Code: Select all

<sshd>
  <port>22</port>
  <pubkeyauthentication/>
  <private-key/>
  <subsystem/>
  <enable/>
  <tcpforwarding/>
  <challengeresponseauthentication/>
</sshd>

I get the following in the file /etc/ssh/sshd_config:

Code: Select all

nas4free: /# cat /etc/ssh/sshd_config
HostKeyAlgorithms ssh-ed25519,ssh-rsa,ssh-dss
HostKey /var/etc/ssh/ssh_host_rsa_key
SyslogFacility LOCAL3
Protocol 2
UseDNS no
Subsystem sftp /usr/libexec/sftp-server
ChallengeResponseAuthentication yes
Port 22
AllowTcpForwarding yes
PubkeyAuthentication yes
ChallengeResponseAuthentication no
AllowTcpForwarding no

Notice the extra ChallengeResponseAuthentication and AllowTcpForwarding in the end.

If I add Additional Parameters in the WebGUI then I get this:

Code: Select all

nas4free: /# cat /etc/ssh/sshd_config
HostKeyAlgorithms ssh-ed25519,ssh-rsa,ssh-dss
HostKey /var/etc/ssh/ssh_host_rsa_key
SyslogFacility LOCAL3
Protocol 2
UseDNS no
Subsystem sftp /usr/libexec/sftp-server
ChallengeResponseAuthentication yes
Port 22
AllowTcpForwarding yes
PubkeyAuthentication yes
# Additional Parameters
ChallengeResponseAuthentication no
AllowTcpForwarding no

The additional parameters goes before the extra ChallengeResponseAuthentication and AllowTcpForwarding.

The problem is in /etc/rc.d/sshd. When using -m "//sshd" I thinkt that it will also include the path /sshd, and for that path the expressions -i "count(challengeresponseauthentication) = 0" and -i "count(tcpforwarding) = 0" are always true, so that is where the two extra 'no' lines at the end come from.

The solution is to move the check for those two to outside the -m "//sshd":

Code: Select all

/usr/local/bin/xml sel -t \
  -o "HostKeyAlgorithms ssh-ed25519,ssh-rsa,ssh-dss" -n \
  -o "HostKey /var/etc/ssh/ssh_host_rsa_key" -n \
  -o "SyslogFacility LOCAL3" -n \
  -o "Protocol 2" -n \
  -o "UseDNS no" -n \
  -i "string-length(//sshd/subsystem) > 0" -v "concat('Subsystem ', //sshd/subsystem)" -n -b \
  -i "string-length(//sshd/subsystem) = 0" -v "concat('Subsystem ', 'sftp /usr/libexec/sftp-server')" -n -b \
  -i "count(//sshd/challengeresponseauthentication) = 0" -o "ChallengeResponseAuthentication no" -n -b \
  -i "count(//sshd/challengeresponseauthentication) > 0" -o "ChallengeResponseAuthentication yes" -n -b \
  -i "count(//sshd/tcpforwarding) = 0" -o "AllowTcpForwarding no" -n -b \
  -i "count(//sshd/tcpforwarding) > 0" -o "AllowTcpForwarding yes" -n -b \
  -m "//sshd" \
    -i "count(port) > 0" -v "concat('Port ',port)" -n -b \
    -i "count(permitrootlogin) > 0" -o "PermitRootLogin yes" -n -b \
    -i "count(compression) > 0" -o "Compression yes" -n -b \
    -i "count(passwordauthentication) > 0" -o "PasswordAuthentication yes" -n -b \
    -i "count(pubkeyauthentication) > 0" -o "PubkeyAuthentication yes" -n -b \
    -m "auxparam" \
      -v "." -n \
    -b \
  -b \
${configxml_file} | /usr/local/bin/xml unesc > ${sshd_config}

[doktornotor]

IPv6 is completely no-op without manual messing. Even with static, it just doesn't work because router advertisement is disabled.

Needed to add

Code: Select all

/sbin/ifconfig bge0 inet6 accept_rtadv

to System > Advanced > Command Scripts (PreInit), plus

Code: Select all

net.inet6.ip6.accept_rtadv = 1

to System > Advanced > sysctl.conf

IPv6 DNS servers also not assigned (did not bother with manual hacks around this).

On another IPv6 note - the GUI just cannot be accessed via IPv6, due to the absolutely horrible "Hosts Allow" misfeature. Kindly put a checkbox there to disable this super-annoying nonsense (there is firewall for those who care).

Add new framework to LAGG

What does it mean ?
Better compatibility with lagg in FreeBSD and nas4free ?

No idea, all I get with lagg are kernel panics - https://sourceforge.net/p/nas4free/bugs/323/

[ms49434]

Hi
The sshd_config file is created incorrectly, the ChallengeResponseAuthentication and AllowTcpForwarding is added twice, where the second one is always no.

I cannot replicate your issue. Please check your config.xml (manually) if there is more than one <sshd></sshd> section.

tested with the following configuration:
config.xml

Code: Select all

        <sshd>
                <port>22</port>
                <passwordauthentication/>
                <pubkeyauthentication/>
                <permitrootlogin/>
                <private-key/>
                <subsystem/>
                <challengeresponseauthentication/>
                <enable/>
                <auxparam>AddressFamily any</auxparam>
        </sshd>

sshd_config

Code: Select all

nas4free: ~# cat /var/etc/ssh/sshd_config
HostKeyAlgorithms ssh-ed25519,ssh-rsa,ssh-dss
HostKey /var/etc/ssh/ssh_host_rsa_key
SyslogFacility LOCAL3
Protocol 2
UseDNS no
Subsystem sftp /usr/libexec/sftp-server
ChallengeResponseAuthentication yes
Port 22
PermitRootLogin yes
AllowTcpForwarding no
PasswordAuthentication yes
PubkeyAuthentication yes
AddressFamily any

[doktornotor]

Some more feedback: the serial console is not usable, the checkbox in System - Advanced does nothing useful plus is missing quite a bunch of required settings. The missing bits are:

/etc/sysctl.conf

Code: Select all

boot_multicons	= YES
boot_serial = YES
console = comconsole,vidconsole
comconsole_port = 0x3f8
comconsole_speed = 115200
hint.uart.0.flags = 0x10

(the last 3 obviously being HW/configuration specific).

Only after adding those I'm actually able to get boot output and console on COM1. (Tested with HP Microserver gen8 and the serial-over-lan.)

Code: Select all

# dmesg | grep uart
uart0: <Non-standard ns8250 class UART with FIFOs> at port 0x3f8 irq 4 flags 0x10 on isa0
uart0: console (115200,n,8,1)

Code: Select all

$ ipmitool -I lanplus -H ilo.example.com -U admin sol activate
Password:
[SOL Session operational.  Use ~? for help]

    NAS4Free: 11.0.0.4 (revision 4249)
    Compiled: Fri May  5 18:27:04 CEST 2017 Platform: x64-embedded
    Copyright (c) 2012-2017 The NAS4Free Project. All Rights Reserved.
    Visit: https://www.nas4free.org for system updates!

        WebGUI Address: https://10.0.0.200

        LAN Network IPv4 Address: 10.0.0.200
        LAN Network IPv6 Address: fe80::2fd:45ff:fefc:f144%bge0
        LAN Network Interface:    bge0

Console Menu
------------
1) Configure Network Interfaces    10) Configure Hosts Allow for WebGUI
2) Configure Network IP Address    11) Restart WebGUI
3) Reset WebGUI Password           12) Restart WebGUI, force HTTP on port 80
4) Reset to Factory Defaults       20) Console Keyboard Map
5) Ping Host
6) Shell
7) Reboot Server
8) Shutdown Server

Enter a number:

[carloskar]

Hi
The sshd_config file is created incorrectly, the ChallengeResponseAuthentication and AllowTcpForwarding is added twice, where the second one is always no.

I cannot replicate your issue. Please check your config.xml (manually) if there is more than one <sshd></sshd> section.

I found the extra <sshd/>, if you enable SSH Events in Remote Syslog Server then the original generation of sshd_config adds the two extra lines.

[JoseMR]

...
Is there something terribly strage with nas4free rootonzfs pool?

Hello MikeMac, I has been using RootOnZFS way before beta test script was released, I have never experienced strange issues other than drives with old GPT/ZFS metadata, while a fresh installation should take care of cleaning it some times I had to quick zero-fill this misbehaving drives.

TIP: After fresh installing NAS4Free RootOnZFS on a single or pair of cleaned-up drives, right after installation go to Tools > Execute Command and execure zpool import -f -a (zpool import -f -a -d /dev/gpt if you use GPT Labeling) then Restore your configuration, there is no need to try to import "zroot" but synchronize.

By the way, I will make several testing on VM and I will fresh install latest NAS4Free RootOnZFS on my own server with a pair of Samsung Fit 32GB USB3 drives just to make sure all is working as expected.

EDIT: While NAS4Free RootOnZFS is working fine on my initial test with a pair of VBox drives previously used in Embedded(UFS) then Zroot, I've noticed strange funny mess information under Disks > Management > HDD Management even after cleaning metadata with a script:

mess.PNG

I will repeat same test after zero-filling this 2 drives and update results.

UPDATE: After zero filling this misbehaving drives with mixed old GPT/ZFS metadata(supposedly cleaned up during install), everything went normally as expected

mess_fixed.PNG

I will perform more testing about different metadata cleanup techniques, since zero-fill maybe overkill and rather slow, though this is a one time proper solution.

Regards

[ms49434]

Hi
The sshd_config file is created incorrectly, the ChallengeResponseAuthentication and AllowTcpForwarding is added twice, where the second one is always no.

I cannot replicate your issue. Please check your config.xml (manually) if there is more than one <sshd></sshd> section.

I found the extra <sshd/>, if you enable SSH Events in Remote Syslog Server then the original generation of sshd_config adds the two extra lines.

the issue has been fixed today with commit r4257.

[MikeMac]

I has been using RootOnZFS way before beta test script was released, I have never experienced strange issues

Thank you, JoseMR!
Probably due to my English, I have not explained understandable, sorry.

1) There is everyting OK with RootOnZFS instalation - it works. And could I use this opportunity and thank you personally for your contribution!

2) But I have made small test. The machine (with both RootOnZFS system pool and data pool) was booted from USB flash with same (4249) version na4free, but embedded. Then I try to import data pool, one had been imported OK

But then I try to import zroot pool, I have a terrible problem. Import itself provides no error messages, but any zpool command after such import attempt provides message about zfs internal error and system becomes unstable.

3) I understand such test is not normal, but IMHO is not forbidden. My idea was about possibility of information extract from ZFS root pool if something goes wrong and system could not start from this pool. I had prepared, that pool could be imported. I could understand, if system will refuse to import zroot pool. But I am afraid zfs lib internal error and unstable system is something wrong

4) As I had found later and wrote in update, it is not issue of 4249 and even nas4free itself, problem could be reproduced with FreeBSD install Live CD is on virtual machine. So may be my message here is offtop, but I had not know it then I initially wrore message, sorry.

[JoseMR]

But then I try to import zroot pool, I have a terrible problem. Import itself provides no error messages, but any zpool command after such import attempt provides message about zfs internal error and system becomes unstable.

Hello MikeMac, I understand your specific situation now on booting LiveMedia/Embedded and recover zroot just in case something goes wrong.

However, this needs to be performed quite different and I will try to explain to you, you cannot import the "zroot" pool normally, since this will directly try to mount the pool at "/" and will completely mess-up the already mounted root file system, you need to import "zroot" with alternate root directory to avoid such behavior with the following example:

Code: Select all

zpool import -f -R /mnt/zroot zroot

then you can check datasets/snaps from the "zroot" pool with "zfs list" or "zfs list -t snapshot".

Regards

[MikeMac]

you cannot import the "zroot" pool normally, since this will directly try to mount the pool at "/" and will completely mess-up the already mounted root file system, you need to import "zroot" with alternate root directory to avoid such behavior with the following example:

Code: Select all

zpool import -f -R /mnt/zroot zroot

Thank you very much for simple explanation!
I have tested - it works.

[nicolap8]

My message of yesterday disappearead, why?
Reposting...

Context: a RAID controller that exposes only the /dev/pass device but not also the /dev/da device (maybe it's a driver bug, not important here). The significat thing is that some disk has ONLY the passthrough device.

The "Disk -> Management" windows when doing an "Import" or an "Add" show and permitt to use also the /dev/pass devices. But these devices are not disks, are usefull only to access configuration, smart, etc and not data.
Trying to use (eg. format) one of these "disk" lead to insuccess.

So the web interface must not uses these devices as disks.

[JoseMR]

My message of yesterday disappearead, why?
Reposting...

Context: a RAID controller that exposes only the /dev/pass device but not also the /dev/da device (maybe it's a driver bug, not important here). The significat thing is that some disk has ONLY the passthrough device.

The "Disk -> Management" windows when doing an "Import" or an "Add" show and permitt to use also the /dev/pass devices. But these devices are not disks, are usefull only to access configuration, smart, etc and not data.
Trying to use (eg. format) one of these "disk" lead to insuccess.

So the web interface must not uses these devices as disks.

While more hardware information in this regards is required, I can tell you that NAS4Free plays best with onboard controllers and HBA's, though some RAID controllers can be flashed to IT mode or simply set to JBOD/RAID-0 mode under the controller utility/BIOS and works ok, however few RAID controllers don't play well and just lead to some odd issues and further problems which can lead to lost of data.

Please check for your specific RAID controller compatibility under the FreeBSD 11.0-RELEASE Hardware Notes.

Regards

[nicolap8]

While more hardware information in this regards is required, I can tell you that NAS4Free plays best with onboard controllers and HBA's, though some RAID controllers can be flashed to IT mode or simply set to JBOD/RAID-0 mode under the controller utility/BIOS and works ok, however few RAID controllers don't play well and just lead to some odd issues and further problems which can lead to lost of data.

Please check for your specific RAID controller compatibility under the FreeBSD 11.0-RELEASE Hardware Notes.

Regards

Ever the same USELESS thing: don't use RAID, use ZFS, blah blah blah...
I not asked for help or suggestions but only REPORTED A BUG.

Of course my controller is in the complatible list but, like all pro know it doesn't signify nothing!

The problem is simple: if in the "camcontrol devlist" exists a disk with only /dev/pass but without /dev/da the web interface has to ignore it or (better) report that there is a problem!