Beta-3 Requests

limetech · August 16, 2010

Beta-3 is ready to post. Here's the current change log:

Changes from 5.0-beta2 to 5.0-beta3
-----------------------------------

Bug Fixes:
- webGui: fix problem using 'default.png' icon for pages without an explicit icon file
- webGui: fix CIFS settings not showing up for Flash share.

Other:
- webGui: treat 'flash' as device on Main page
- kernel: add 'asus_atk0110' kernel module
- kernel: change other supported hwmon components to modules instead of built-in
- kernel: make 'coretemp' a module instead of built-in
- Include 'infozip' package for zip/unzip support.

There is an issue with getting status from the md-driver which I need to fix & will also probably release 4.5.7 as well with this fix.

In order to update to latest packages for AFP support (and some other goodies), I want to upgrade to latest Slack release (13.1) and perhaps upgrade to 2.6.35 linux kernel.

Reason for this post is to let you request things you need for -beta3 that I can throw in before embarking on above 'major' changes. I'd like to get -beta3 out this week, preferably by Wednesday, so the window for requests is open for a small time.

WeeboTech · August 16, 2010

Is there any way to have a real /cgi-bin mechanism.

If I were to package up NAGIOS for monitoring unRAID, it compiles C cgi programs for managing the alerts.

I could hook it into lighttpd, I just figure that if emhttp can do real cgi requests a monitoring environment would not depend on another webserver.

WeeboTech · August 16, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

BRiT · August 16, 2010

Beta-3 is ready to post.

In order to update to latest packages for AFP support (and some other goodies), I want to upgrade to latest Slack release (13.1) and perhaps upgrade to 2.6.35 linux kernel.

Add another +1 to ext3 and ext4 support.

As an FYI, I have unRAID 5.0 beta 2 running on kernel 2.6.35.2 without issues on a 3 drive setup. I did have to do 1 small patch to md.c to make it compile, that being 'blk_queue_max_sectors' to 'blk_queue_max_hw_sectors'.

--- /cache/.root/unraid/50b2/usr/src/linux/drivers/md/md.c      2010-07-19 05:16:45.000000000 -0400
+++ /usr/src/linux/drivers/md/md.c      2010-08-14 00:47:05.000000000 -0400
@@ -1232,7 +1232,7 @@
                        gd->queue->queue_lock = &mddev->queue_lock;
                        gd->queue->queuedata = mddev;
                        blk_queue_make_request(gd->queue, md_make_request);
-                       blk_queue_max_sectors(gd->queue, 256);  /* 256 sectors => 128K */
+                       blk_queue_max_hw_sectors(gd->queue, 256);  /* 256 sectors => 128K */
                        gd->queue->backing_dev_info.ra_pages =
                                (512 * 1024) / PAGE_SIZE;

queeg · August 16, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

Add another +1 for me.

NAS · August 16, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

Add another +1 for me.

And +1 from me. Would have saved me no end of faffing recently. Some sort of simple encrypted fs support for physically moving data using USB HDDs semi securely between NAS boxes would be nice if it can be done trivially. But ext3/4 alone would be a huge step alone.

gfjardim · August 16, 2010

The tun/tap device driver and the 802.d bridging module for OpenVPN remote access plugin:

[ * ] Networking support --->

Networking options --->

<M> 802.1d Ethernet Bridging

Device Drivers --->

[ * ] Network device support --->

<M> Universal TUN/TAP device driver support

The K8 Powernow driver for AMD power management:

Power management and ACPI options --->

CPU Frequency scaling --->

[ * ] CPU Frequency scaling

Default CPUFreq governor (ondemand) --->

<M> AMD Opteron/Athlon64 PowerNow!

<M> ACPI Processor P-States driver

BRiT · August 16, 2010

The K8 Powernow driver for AMD power management:

Power management and ACPI options --->

CPU Frequency scaling --->

[ * ] CPU Frequency scaling

Default CPUFreq governor (ondemand) --->

<M> AMD Opteron/Athlon64 PowerNow!

<M> ACPI Processor P-States driver

Add +1 for the ACPI Processor P-States driver. I'm not sure if it's only for Linux 2.6.35 series of kernels.

The ACPI Processor Performance States (CONFIG_X86_ACPI_CPUFREQ). It is utilized for Intel i3/i5/i7 power management.

Power management and ACPI options --->

CPU Frequency scaling --->

[ * ] CPU Frequency scaling

<M> ACPI Processor P-States driver

This driver adds a CPUFreq driver which utilizes the ACPI

Processor Performance States.

This driver also supports Intel Enhanced Speedstep.

To compile this driver as a module, choose M here: the

module will be called acpi-cpufreq.

For details, take a look at <file:Documentation/cpu-freq/>.

If in doubt, say N.

Symbol: X86_ACPI_CPUFREQ [=m]

Prompt: ACPI Processor P-States driver

Defined at arch/x86/kernel/cpu/cpufreq/Kconfig:27

Depends on: CPU_FREQ [=y] && ACPI_PROCESSOR [=m]

Location:

-> Power management and ACPI options

-> CPU Frequency scaling

-> CPU Frequency scaling (CPU_FREQ [=y])

Selects: CPU_FREQ_TABLE [=m]

bubbaQ · August 16, 2010

+1 ext3...

Please add all the frequency scaling modules... (no need for debug or stats).

Current Intel SpeedStep options are all for mobile chipsets, and not needed.

[*] CPU Frequency scaling

[ ] Enable CPUfreq debugging

< > CPU frequency translation statistics

Default CPUFreq governor (performance) --->

-*- 'performance' governor

<M> 'powersave' governor

<M> 'userspace' governor for userspace frequency scaling

<*> 'ondemand' cpufreq policy governor

<M> 'conservative' cpufreq governor

*** CPUFreq processor drivers ***

<M> ACPI Processor P-States driver

<M> AMD Mobile K6-2/K6-3 PowerNow!

<M> AMD Mobile Athlon/Duron PowerNow!

<M> AMD Opteron/Athlon64 PowerNow!

[*] ACPI Support

<M> Cyrix MediaGX/NatSemi Geode Suspend Modulation

< > Intel Enhanced SpeedStep (deprecated)

< > Intel Speedstep on ICH-M chipsets (ioport interface)

< > Intel SpeedStep on 440BX/ZX/MX chipsets (SMI interface)

<M> Intel Pentium 4 clock modulation

<M> nVidia nForce2 FSB changing

CD/DVD filesystems support would be nice too.

When you get to big changes, I would like to see some type of native encryption... or at least some hooks for it so a plugin could be made that can do it.

Nice to hear about the move to 13.1 too.

BRiT · August 16, 2010

Current Intel SpeedStep options are all for mobile chipsets, and not needed.

Even if it is needed, from how I read the write ups, the ACPI Processor P-States driver seems to support most (all?) that the deprecated Intel SpeedStep driver did.

bubbaQ · August 17, 2010

Even if it is needed, from how I read the write ups, the ACPI Processor P-States driver seems to support most (all?) that the deprecated Intel SpeedStep driver did.

Yes, but it requires ACPI, and some people disable ACPI for stability reasons. p4clockmod does not need ACPI.

limetech · August 17, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

Also, as part of implementing multiple file system selection, I wanted to implement a "maintenance mode" - this would be a checkbox that would show up when array is stopped, so that, if checked, when you click Start button, array is brought to point where 'md' devices are valid (and writes update parity), but no file systems are mounted, and no network services are started (except for CIFS access to the 'flash' share). The reason for 'maintenance mode' is to let you run 'fsck' and other disk utilities that must run without any file systems mounted.

So... probably this is too much for -beta3?

WeeboTech · August 17, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

First step is kernel support. Yes there have been requests to support disk-by-disk file system format.

Recently there have been requests for the ability to mount foreign (non array) disks.

Including kernel support is step 1, so foreign disks can be mounted.

limetech · August 17, 2010

This sufficient?

[*] CPU Frequency scaling
[ ]   Enable CPUfreq debugging (NEW)
<*>   CPU frequency translation statistics (NEW)
[ ]     CPU frequency translation statistics details (NEW)
Default CPUFreq governor (ondemand)  --->
-*-   'performance' governor
<M>   'powersave' governor
<M>   'userspace' governor for userspace frequency scaling
-*-   'ondemand' cpufreq policy governor
<M>   'conservative' cpufreq governor
*** CPUFreq processor drivers ***
<M>   ACPI Processor P-States driver
<M>   AMD Mobile K6-2/K6-3 PowerNow!
<M>   AMD Mobile Athlon/Duron PowerNow!
<M>   AMD Opteron/Athlon64 PowerNow!
< >   Cyrix MediaGX/NatSemi Geode Suspend Modulation (NEW)
< >   Intel Enhanced SpeedStep (deprecated) (NEW)
< >   Intel Speedstep on ICH-M chipsets (ioport interface) (NEW)
< >   Intel SpeedStep on 440BX/ZX/MX chipsets (SMI interface) (NEW)
<M>   Intel Pentium 4 clock modulation
<M>   nVidia nForce2 FSB changing
< >   Transmeta LongRun (NEW)
< >   VIA Cyrix III Longhaul (NEW)
< >   VIA C7 Enhanced PowerSaver (DANGEROUS) (NEW)
*** shared options ***

WeeboTech · August 17, 2010

Also, as part of implementing multiple file system selection, I wanted to implement a "maintenance mode" - this would be a checkbox that would show up when array is stopped, so that, if checked, when you click Start button, array is brought to point where 'md' devices are valid (and writes update parity), but no file systems are mounted, and no network services are started (except for CIFS access to the 'flash' share). The reason for 'maintenance mode' is to let you run 'fsck' and other disk utilities that must run without any file systems mounted.

So... probably this is too much for -beta3?

I'm OK with testing out a maintenance mode if you want us to check it out.

BRiT · August 17, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

For beta3, I'd be happy enough if unRAID had the kernel modules for ext3 and ext4 available (CONFIG_EXT3_FS, CONFIG_EXT4_FS).

We can worry about using filesystems other than REISERFS for the unRAID MD disks later (beta 4 or later).

Good idea on the maintenance mode / failsafe boot modes. I'm guessing this could also be passed as a boot-time parameter option as well. I know there's some community discussions about such a thing somewhere around here.

I just found my nugget of the day, thanks Limetech for the following snippet:

< > VIA C7 Enhanced PowerSaver (DANGEROUS) (NEW)

limetech · August 17, 2010

Hows this? (POSIX ACL's are necessary for Active Directory support):

<*> Second extended fs support
[*]   Ext2 extended attributes
[*]     Ext2 POSIX Access Control Lists
[ ]     Ext2 Security Labels (NEW)
[ ]   Ext2 execute in place support
<*> Ext3 journalling file system support
[ ]   Default to 'data=ordered' in ext3 (NEW)
[*]   Ext3 extended attributes (NEW)
[*]     Ext3 POSIX Access Control Lists
[ ]     Ext3 Security Labels (NEW)
<*> The Extended 4 (ext4) filesystem
[*]   Ext4 extended attributes (NEW)
[*]     Ext4 POSIX Access Control Lists
[ ]     Ext4 Security Labels (NEW)
[ ]   EXT4 debugging support (NEW)

WeeboTech · August 17, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

For beta3, I'd be happy enough if unRAID had the kernel modules for ext3 and ext4 available (CONFIG_EXT3_FS, CONFIG_EXT4_FS).

We can worry about using filesystems other than REISERFS for the unRAID MD disks later (beta 4 or later).

Good idea on the maintenance mode / failsafe boot modes. I'm guessing this could also be passed as a boot-time parameter option as well. I know there's some community discussions about such a thing somewhere around here.

We were discussing "safe" as a boot time option which could be checked with

if grep -qi safe /proc/cmdline

This would skip the installation of packages in /boot/extra.

Now if emhttp can see this and go into it's maintain or safe mode that would be a big help.

Joe L. · August 17, 2010

There was a request for ext3 and ext4.

I'm adding a +1 to it also.

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

Also, as part of implementing multiple file system selection, I wanted to implement a "maintenance mode" - this would be a checkbox that would show up when array is stopped, so that, if checked, when you click Start button, array is brought to point where 'md' devices are valid (and writes update parity), but no file systems are mounted, and no network services are started (except for CIFS access to the 'flash' share). The reason for 'maintenance mode' is to let you run 'fsck' and other disk utilities that must run without any file systems mounted.

So... probably this is too much for -beta3?

The ability to use alternate file-systems in the unRAID protected array is a huge change. I do not think that is what is being asked for beta3.

I think what is being requested is only the inclusion of the drivers, so that ext4 and ext4 drives from other servers can be mounted by the technically inclined for data migration. I do not think anybody expects you to add the ability to format, use, and recognize as NOT un-formatted those alternate file-systems for the protected array in the 5.0 series. It would be nice for the 5.1 series though..., once most of the graphical stuff settles down.

But, now that alternate file systems are being mentioned, is it about time to replace the old (read-only) ntfs driver with the (read-write) ntfs-3g driver. ntfs-3g is very stable, and very nice to be able to migrate data. It has come a looooooong way in the past few years.

http://www.tuxera.com/community/release-history/

limetech · August 17, 2010

Also for EXT3, there's an option, either:

'data=ordered' or

'data=writeback'

Seems to me for unRAID 'data=writeback' is preferred, but if you know anything about this, please chime in.

limetech · August 17, 2010

I think what is being requested is only the inclusion of the drivers, so that ext4 and ext4 drives from other servers can be mounted by the technically inclined for data migration. I do not think anybody expects you to add the ability to format, use, and recognize as NOT un-formatted those alternate file-systems for the protected array in the 5.0 series. It would be nice for the 5.1 series though..., once most of the graphical stuff settles down.

Ok, that's what I thought & how it will be included in -beta3. I'm also a bit nervous about how large the resulting kernel is going to be... we'll see.

But, now that alternate file systems are being mentioned, is it about time to replace the old (read-only) ntfs driver with the (read-write) ntfs-3g driver. ntfs-3g is very stable, and very nice to be able to migrate data. It has come a looooooong way in the past few years.

http://www.tuxera.com/community/release-history/

Do you know if ntfs-3g now supports ACL's? (Again necessary for Active Directory, which you would think would be supported by a microsoft file system clone, but last I checked, wasn't there.)

limetech · August 17, 2010

There is an issue with getting status from the md-driver which I need to fix & will also probably release 4.5.7 as well with this fix.

In order to update to latest packages for AFP support (and some other goodies),

I want to upgrade to latest Slack release (13.1) and perhaps upgrade to 2.6.35 linux kernel.

I hope you mean to do that for both 4.5.7 and 5.0-beta.

Sorry, 4.x is now in 'maintenance' stage - no new features.

Joe L. · August 17, 2010

Do you know if ntfs-3g now supports ACL's? (Again necessary for Active Directory, which you would think would be supported by a microsoft file system clone, but last I checked, wasn't there.)

Added last year, several versions ago.

STABLE Version 2009.11.14 (November 15, 2009)

* New: Full file ownership and permissions support. The ownership and permissions of files are interoperable with Windows and conforms to the POSIX rules.

* New: POSIX ACL support. The ACLs are an extension to ownership and permissions as defined in POSIX.

* New: Optional Windows-type inheritance of ACLs.

* New: Full extended attributes support for all namespaces: user, system, security and trusted.

The newest stable version even transparently supports read/write compressed file-systems.

bubbaQ · August 17, 2010

This is probably not something I can implement in just a couple days, since I think what you want is to be able to specify on a disk-by-disk basis what file system to format each disk as, correct?

My request was *not* for unRAID array disks to use ext3/4... just to have support in the kernel since I often mount drives and dd images from other systems to copy their contents to unRAID... and they are often ext3.

bubbaQ · August 17, 2010

This sufficient?

Sweet!

I completely understand if this makes the kernel too big... if the modules are big, but the kernal is not and you need to shrink the initramfs, perhaps you can post the modules separately for plugin devs to install selectively if needed.

Beta-3 Requests

Recommended Posts

Link to comment

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment