Re: OS6 now works on x86 Legacy WARNING: NO NTGR SUPPORT!

Skywalker · ‎2013-07-31

The iSCSI stack is exactly the same as 4.2.

mangrove · ‎2013-07-31

Thanks, Skywalker.

arbert · ‎2013-08-04

Just installed OS6 on my 4200 to do some testing to see if I trusted it for production use.

If I use RADAR on the initial configuration, it will use all 12 bays for a volume. However, once built, the GUI only displays 4 drives and I can't create or work with any volumes outside of the first 4 bays. Any ideas?

Thanks

beisser1 · ‎2013-08-04

thats because the chipset for channels 5-12 (lsi sas controller) is not supported.
channels 1-4 are handled by the intel sata controller and are supported.

i guess you are out of luck on that one. your best chance of success is with the desktop models. the 12 bay rackmount units will probably never work.

arbert · ‎2013-08-04

Ya, that sucks. I see that burried back in one of the previous posts. How can I get back to the old firmware now--anything I upload is errored as "not a valid NASOS image" and, of course, factory default just resets to a default 6.0.8.

mdgm-ntgr · ‎2013-08-04

Sent you a PM with a link to a modified firmware image which will allow you to downgrade to 4.2.x. I haven't made one for 4.2.x yet but I did make one for 4.2.23. If you want to make one yourself you take a 4.2.x image and modify the header to appear as if it is a 6.0.x image. So it was made the same way as the image you used to update to 6.0.8

DariusSki · ‎2013-08-05

I have installed 6.0.8 on Ultra4 and everything seems to be fine, front page reporting is broken, but other than that everything is fine. Even iSCSI is working great ( from windows 7 clients using mutual CHAP).

I have a bit "advanced" problems - please note that they should not stop anyone from using OS6 on older gear, they are just plea for advanced help:

1) I don't seem to be able to stop the NAS from "calling" home, i ran apt-get purge on some packages as recommended earlier in this thread, but i still get this:
root@poligonas:~# netstat -tan
tcp 0 0 192.168.1.72:35191 206.16.42.179:443 ESTABLISHED

Anyone got clue how to get rid of this stuff?

2) Power efficiency of OS6, it just seems to be chewing a few watts more than 4.2.24. While i don't really care about disk turning off, i decided to dig deeper and didn't like what i've found:
After running
pidstat -t -w
i saw this:
12:52:11 AM 1673 - 1.00 0.00 readynasd
12:52:11 AM - 1673 1.00 0.00 |__readynasd
12:52:11 AM - 1734 100.00 1.00 |__readynasd
12:52:11 AM - 1735 100.00 0.00 |__readynasd
12:52:11 AM - 1736 100.00 0.00 |__readynasd
12:52:11 AM - 1737 100.00 0.00 |__readynasd
12:52:11 AM - 1738 100.00 2.00 |__readynasd
12:52:11 AM - 1740 100.00 0.00 |__readynasd
12:52:11 AM - 1741 100.00 0.00 |__readynasd
12:52:11 AM - 1749 100.00 0.00 |__readynasd
12:52:11 AM - 1750 1.00 0.00 |__readynasd
12:52:11 AM - 1761 1.00 0.00 |__readynasd
12:52:11 AM - 1938 1.00 0.00 |__readynasd

Yup, you see it right, it is readynasd waking your CPU 800 times per second and not letting it go to deeper sleep states.
And unfortunately those threads are not doing anything important either, they are using technique that has fallen out of favor in any power use aware programming for years now:
strace -r -p 1734
Process 1734 attached - interrupt to quit
0.000000 restart_syscall(<... resuming interrupted call ...>) = 0
0.006898 nanosleep({0, 10000000}, NULL) = 0
0.010366 nanosleep({0, 10000000}, NULL) = 0

Yup, that is nanosleep for 10ms burning extra electricity.

4.2.24 had none of this, looking at vmstat you could see dozens of interrupts/wakeups instead of thousand+ on OS6 box...

Question is - can we get rid of all this madness? I can't dig deeper without gdb and powertop, but maybe some Jedi will have a clue wth is going on with those 8 threads?

3) Anyone got idea why Netgear is using acpi_idle instead of intel_idle? while both are using mwait in idle, intel_idle can go deeper. I wonder if anyone tested custom kernel with intel_idle, is it working?

Skywalker · ‎2013-08-09

DariusSki wrote:
1) I don't seem to be able to stop the NAS from "calling" home, i ran apt-get purge on some packages as recommended earlier in this thread, but i still get this:
root@poligonas:~# netstat -tan
tcp 0 0 192.168.1.72:35191 206.16.42.179:443 ESTABLISHED

Anyone got clue how to get rid of this stuff?

Which process is it?

# apt-get purge busybox-net-tools
# apt-get install net-tools
# netstat -tanp

DariusSki wrote:
2) Power efficiency of OS6, it just seems to be chewing a few watts more than 4.2.24. While i don't really care about disk turning off, i decided to dig deeper and didn't like what i've found:
...
Question is - can we get rid of all this madness? I can't dig deeper without gdb and powertop, but maybe some Jedi will have a clue wth is going on with those 8 threads?

Yes, you should find that the next release behaves much better in this regard. We should have a beta very soon.

DariusSki wrote:
3) Anyone got idea why Netgear is using acpi_idle instead of intel_idle? while both are using mwait in idle, intel_idle can go deeper. I wonder if anyone tested custom kernel with intel_idle, is it working?

Just a simple oversight. intel_idle will be enabled in the next release as well.

DariusSki · ‎2013-08-09

Skywalker wrote:

Which process is it?

tcp 0 0 192.168.1.72:35683 206.16.42.179:443 ESTABLISHED 16086/rcbrokerd

Skywalker wrote:

Yes, you should find that the next release behaves much better in this regard. We should have a beta very soon.

Just an simple oversight. intel_idle will be enabled in the next release as well.

AWESOME, thanks a lot for keeping community updated. Looking forward to these low hanging fruits of power efficiency. OS6 is really shaping up.

mdgm-ntgr · ‎2013-08-10

6.1.1 beta firmware is available if you would like to try it: http://www.readynas.com/forum/viewtopic.php?f=154&t=72282

Iridabike · ‎2013-08-10

Took the plunge and have 6.1.1 beta running on my Pro 6. Mostly all ok. Fans, temps and temp rrd trends still don't work on Pro 6 without some mods/patches.

Have some problems moving the apache port from 80 to 8080 as ports.conf no longer seems to be recognized but other than that all seems good.

Edit: apache2 ports are fine and can be set. I had a typo! Have noticed one problem though: Network config from frontview gives an error.

Edit2: Network trending isn't working in frontview performance monitoring either on a Pro 6.

Skywalker · ‎2013-08-11

DariusSki wrote:
tcp 0 0 192.168.1.72:35683 206.16.42.179:443 ESTABLISHED 16086/rcbrokerd

That's the ReadyCloud service.

systemctl disable rcbrokerd

should take care of it. There's supposed to be a way to disable this from the UI, but it doesn't appear to be straightforward at this point.

DariusSki · ‎2013-08-12

ran


root@poligonas:~# systemctl disable rcbrokerd
rm '/etc/systemd/system/multi-user.target.wants/rcbrokerd.service'

but i can still see the process of "rcbrokerd". Killing it is not working either, it restarts immediately 🙂

Skywalker · ‎2013-08-12

Oh, right, that's the saved state and wouldn't take effect until you reboot. For immediate action, you need to add "systemctl stop rcbrokerd".

Iridabike · ‎2013-08-12

6.1.1 beta on Pro 6 problems so far:

1) Network card is not recognised in frontview - gives an error and shows a blank screen and hence cannot change any network settings
2) Network data does not show in the performance trend
3) Cannot download the logs

Plus the usual problem with temps and fans

Does anyone know if I can switch back to 6.0.8?

demetris · ‎2013-08-13

So some good news update
It seems sabnzb is added to the utility list of netgear yay!
apt-get update
apt-get install sabnzbd
will install: libcgi-pm-perl python-yenc libcgi-pm-perl python-yenc sabnzbd
Version? sabnzbd_0.7.13-0rnx1_all.deb Awesome!

Get:2 http://apt.readynas.com/packages/readynasos/ 6.0.8/apps sabnzbd all 0.7.13-0rnx1 [2695 kB]
Before you update go to sabnzb config then select General tab.
At SABnzbd Port change to 2708 and save changes then restart sabnzbd.
After Official Netgear install it will show up in apps panel and be able to launch it from there

mdgm-ntgr · ‎2013-08-13

Just installed 6.1.1 beta on my Pro 2.

It now shows temps and fan speed which I think might be accurate? Though it looks like there still isn't any fan control. The Pro 2 has a fan speed of close to (but under) 2000 RPM, CPU temp of 19 degrees celsius and system temp of 25 degrees celsius.

tkapin · ‎2013-08-14

mdgm ... yes, you are correct .. , but this build looks very promising .. for us 🙂
only one thing was need from my side, re-configure sensors, because i was get dead fan error message .. my fans was set too low 🙂

Iridabike · ‎2013-08-14

Installed T1826 on my Pro 6:

Network configuration is now fixed but alas no temps and fans for me 😞

super_poussin · ‎2013-08-14

arnomc wrote:
About Dynamic DNS (which i need), I have seen this 4-5 year old post, but would it work on a OS 6 system ?
viewtopic.php?p=110351
Or is there any other solution (linux package ddns?) because RN6 is a modified debian wheezy/sid, like this one for exple :
http://wiki.debian.org/DDNS or http://packages.debian.org/stable/ddns3-client
(please note: I am not familiar with any linux distrib)
Anybody here could give me a hint on this ?
It would be VERY much appreciated

you should see a ddclient under free apps on frontview

drfrogsplat · ‎2013-08-15

If I have an existing Ultra 4 set up with RAIDiator 4.x, is it possible to put a whole new set of drives in (leaving my old ones as they were) to test out 6.x, and then if/when things go wrong just swap the 4.x disks back in and continue running?

Or will doing the firmware update + factory reset alter the built-in flash (or anything else?) that would prevent me from running the set of 4.x drives after having run 6.x on a different set of drives?

---

if that's not clear... currently running Ultra 4 with 4.2.23 and 3 drives (A, B, C).. I have spare drives (D, E). I'd like to
- remove A, B, C
- install D, E
- factory reset / upgrade to 6.x / factory reset (as part of the upgrade requirements)
- try out 6.x until something breaks (or i decide to commit to it)
- remove D, E
- put A, B, C back in
- boot back into 4.x as if nothing had ever happened...

StephenB · ‎2013-08-16

drfrogsplat wrote:
...
if that's not clear... currently running Ultra 4 with 4.2.23 and 3 drives (A, B, C).. I have spare drives (D, E). I'd like to
- remove A, B, C
- install D, E
- factory reset / upgrade to 6.x / factory reset (as part of the upgrade requirements)
- try out 6.x until something breaks (or i decide to commit to it)
- remove D, E
- put A, B, C back in
- boot back into 4.x as if nothing had ever happened...

This should work, as long as the NAS is powered down when you swap the drives.

platini76 · ‎2013-08-26

I dont' understand if there are advantages to upgrade or only to test it...
i'm very undecided if upgrde my or not...

StephenB · ‎2013-08-27

platini76 wrote:
I dont' understand if there are advantages to upgrade or only to test it...
i'm very undecided if upgrde my or not...

If you are happy with the features of your pro, than stick with 4.2.24. Netgear will not support you if you switch to OS6.

platini76 · ‎2013-08-27

i'm in italy, netgear don't support me in any case...:(

what I have not understand is about the features... what are the real advantages? which difference in features?