Wifi connection issues with Intel?

[Richard Shaw]

I looked though the last update and nothing jumped out at me...  Tried
journalctl --this-boot and paged through it, nothing jumped out at me.

I've also tried all 3 installed kernels with the same result: Flakey
wireless

When I bring up the connection details (which takes to many clicks/menus) I
see connection speeds of 1, 18, 24, or 54Mbit but this is a n device and
has connected reliably at 108Mbit before now and the AP is less than 20
feet away.

Sometimes connections get stuck at "resolving host" and other times it
works.

Anyone seeing the same issue?

Thanks,
Richard

[Richard Shaw]

The problem seems to have at least partially resolved itself. She now has
good enough connectivity to use the web and connect to the printer on my
desktop like before but the connection speed is still maxing out at 54Mbit
when it used to connect reliably at 108Mbit (still connected at g instead
of n). No hardware has changed nor has the AP firmware been updated so the
only variable here is the Fedora installation....

Richard

[poma]

On 10.06.2014 14:52, Richard Shaw wrote:
> The problem seems to have at least partially resolved itself. She now has
> good enough connectivity to use the web and connect to the printer on my
> desktop like before but the connection speed is still maxing out at 54Mbit
> when it used to connect reliably at 108Mbit (still connected at g instead
> of n). No hardware has changed nor has the AP firmware been updated so the
> only variable here is the Fedora installation....
>
> Richard

Firmware, kernel module or userspace? :)
Did you mention an exact hardware in question?
BTW how are you sure RF environment ain't changed?

poma

[Richard Shaw]

On Tue, Jun 10, 2014 at 6:09 PM, poma  wrote:
> Firmware, kernel module or userspace? :)

# uname -r
3.14.5-200.fc20.x86_64

No clue on firmware, I'm not exactly sure how to check which is being
loaded...

Did you mention an exact hardware in question?
>

02:00.0 Network controller: Intel Corporation Centrino Advanced-N 6235 (rev
24)
Subsystem: Intel Corporation Centrino Advanced-N 6235 AGN
Flags: bus master, fast devsel, latency 0, IRQ 48
Memory at f0200000 (64-bit, non-prefetchable) [size=8K]
Capabilities: [c8] Power Management version 3
Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [e0] Express Endpoint, MSI 00
Capabilities: [100] Advanced Error Reporting
Capabilities: [140] Device Serial Number c8-f7-33-ff-ff-f4-2d-ba
Kernel driver in use: iwlwifi
Kernel modules: iwlwifi

BTW how are you sure RF environment ain't changed?
>

Nothing new has come in or left the house since the problem...

Thanks,
Richard

[poma]

On 11.06.2014 04:34, Richard Shaw wrote:
> On Tue, Jun 10, 2014 at 6:09 PM, poma  wrote:
> > Firmware, kernel module or userspace? :)
>
> # uname -r
> 3.14.5-200.fc20.x86_64
>
> No clue on firmware, I'm not exactly sure how to check which is being
> loaded...
>
> Did you mention an exact hardware in question?
> >
>
> 02:00.0 Network controller: Intel Corporation Centrino Advanced-N 6235 (rev
> 24)
> Subsystem: Intel Corporation Centrino Advanced-N 6235 AGN
> Flags: bus master, fast devsel, latency 0, IRQ 48
> Memory at f0200000 (64-bit, non-prefetchable) [size=8K]
> Capabilities: [c8] Power Management version 3
> Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
> Capabilities: [e0] Express Endpoint, MSI 00
> Capabilities: [100] Advanced Error Reporting
> Capabilities: [140] Device Serial Number c8-f7-33-ff-ff-f4-2d-ba
> Kernel driver in use: iwlwifi
> Kernel modules: iwlwifi
>
> BTW how are you sure RF environment ain't changed?
> >
>
> Nothing new has come in or left the house since the problem...
>
> Thanks,
> Richard

- KERNEL:
$ curl -s https://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.14.6 | grep iwlwifi

So pick up kernel-3.14.6-200.fc20.x86_64.rpm
http://koji.fedoraproject.org/koji/buildinfo?buildID=531773

You can also try with kernel-core-3.15.0-1.fc21.x86_64.rpm
http://alt.fedoraproject.org/pub/alt/rawhide-kernel-nodebug/x86_64/

- FIRMWARE:
$ modinfo -F firmware iwlwifi

# yum --enablerepo updates-testing install iwl6000g2b-firmware

$ rpm -qf /usr/lib/firmware/iwlwifi-6000g2b-6.ucode
iwl6000g2b-firmware-17.168.5.2-38.fc20.noarch

According to
http://wireless.kernel.org/en/users/Drivers/iwlwifi#Firmware
Intel® Centrino® Advanced-N 6235 goes with iwlwifi-6000g2b-ucode-18.168.6.1.tgz
i.e. iwlwifi-6000g2b-6.ucode
also stated within firmware itself:
https://git.kernel.org/cgit/linux/kernel/git/firmware/linux-firmware.git/tre...
"IWL.6000g2b.fw.v18.168.6.1.build.0"

However in Fedora it is packaged as "17" i.e. iwl6000g2b-firmware-17.168.5.2-38.fc20.noarch

However checksums are equal:
   $ sha1sum /usr/lib/firmware/iwlwifi-6000g2b-6.ucode \
> /tmp/iwlwifi-6000g2b-ucode-18.168.6.1/iwlwifi-6000g2b-6.ucode

0c5dbfed5f50ef39b56827651f572bea25bfc27e  /usr/lib/firmware/iwlwifi-6000g2b-6.ucode
   0c5dbfed5f50ef39b56827651f572bea25bfc27e  /tmp/iwlwifi-6000g2b-ucode-18.168.6.1/iwlwifi-6000g2b-6.ucode
So you are good to go with "17".

Josh, do these two firmwares should be in their own packages as "18"?

Intel® Centrino® Advanced-N 6235 - iwlwifi-6000g2b-ucode-18.168.6.1
i.e. iwlwifi-6000g2b-6.ucode - "IWL.6000g2b fw v18.168.6.1 build 0"
in iwl6000g2b-firmware-18.168.6.1-1.f$releasever.noarch.rpm

and

Intel® Centrino® Advanced-N 6205 - iwlwifi-6000g2a-ucode-18.168.6.1
i.e. iwlwifi-6000g2a-6.ucode - "IWL.6000g2a fw v18.168.6.1 build 0"
in iwl6000g2a-firmware-18.168.6.1-1.f$releasever.noarch.rpm

poma

[poma]

On 11.06.2014 10:18, poma wrote:
> On 11.06.2014 04:34, Richard Shaw wrote:
> > On Tue, Jun 10, 2014 at 6:09 PM, poma  wrote:
> > > Firmware, kernel module or userspace? :)
> >
> > # uname -r
> > 3.14.5-200.fc20.x86_64
> >
> > No clue on firmware, I'm not exactly sure how to check which is being
> > loaded...
> >
> > Did you mention an exact hardware in question?
> > >
> >
> > 02:00.0 Network controller: Intel Corporation Centrino Advanced-N 6235 (rev
> > 24)
> > Subsystem: Intel Corporation Centrino Advanced-N 6235 AGN
> > Flags: bus master, fast devsel, latency 0, IRQ 48
> > Memory at f0200000 (64-bit, non-prefetchable) [size=8K]
> > Capabilities: [c8] Power Management version 3
> > Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
> > Capabilities: [e0] Express Endpoint, MSI 00
> > Capabilities: [100] Advanced Error Reporting
> > Capabilities: [140] Device Serial Number c8-f7-33-ff-ff-f4-2d-ba
> > Kernel driver in use: iwlwifi
> > Kernel modules: iwlwifi
> >
> > BTW how are you sure RF environment ain't changed?
> > >
> >
> > Nothing new has come in or left the house since the problem...
> >
> > Thanks,
> > Richard
> KERNEL:
>
> $ curl -s https://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.14.6 | grep iwlwifi
>
> So pick up kernel-3.14.6-200.fc20.x86_64.rpm
> http://koji.fedoraproject.org/koji/buildinfo?buildID=531773
>
> You can also try with kernel-core-3.15.0-1.fc21.x86_64.rpm
> http://alt.fedoraproject.org/pub/alt/rawhide-kernel-nodebug/x86_64/
> FIRMWARE:
>
> $ modinfo -F firmware iwlwifi
>
> # yum --enablerepo updates-testing install iwl6000g2b-firmware
>
> $ rpm -qf /usr/lib/firmware/iwlwifi-6000g2b-6.ucode
> iwl6000g2b-firmware-17.168.5.2-38.fc20.noarch
>
> According to
> http://wireless.kernel.org/en/users/Drivers/iwlwifi#Firmware
> Intel® Centrino® Advanced-N 6235 goes with iwlwifi-6000g2b-ucode-18.168.6.1.tgz
> i.e. iwlwifi-6000g2b-6.ucode
> also stated within firmware itself:
> https://git.kernel.org/cgit/linux/kernel/git/firmware/linux-firmware.git/tre...
> "IWL.6000g2b.fw.v18.168.6.1.build.0"
>
> However in Fedora it is packaged as "17" i.e. iwl6000g2b-firmware-17.168.5.2-38.fc20.noarch
>
> However checksums are equal:
>     $ sha1sum /usr/lib/firmware/iwlwifi-6000g2b-6.ucode \
>     > /tmp/iwlwifi-6000g2b-ucode-18.168.6.1/iwlwifi-6000g2b-6.ucode
>     0c5dbfed5f50ef39b56827651f572bea25bfc27e  /usr/lib/firmware/iwlwifi-6000g2b-6.ucode
>     0c5dbfed5f50ef39b56827651f572bea25bfc27e  /tmp/iwlwifi-6000g2b-ucode-18.168.6.1/iwlwifi-6000g2b-6.ucode
> So you are good to go with "17".
>
> Josh, do these two firmwares should be in their own packages as "18"?
>
> Intel® Centrino® Advanced-N 6235 - iwlwifi-6000g2b-ucode-18.168.6.1
> i.e. iwlwifi-6000g2b-6.ucode - "IWL.6000g2b fw v18.168.6.1 build 0"
> in iwl6000g2b-firmware-18.168.6.1-1.f$releasever.noarch.rpm
>
> and
>
> Intel® Centrino® Advanced-N 6205 - iwlwifi-6000g2a-ucode-18.168.6.1
> i.e. iwlwifi-6000g2a-6.ucode - "IWL.6000g2a fw v18.168.6.1 build 0"
> in iwl6000g2a-firmware-18.168.6.1-1.f$releasever.noarch.rpm

e.g.

- Intel® Centrino® Advanced-N 6235

1. Rename
iwl6000g2b-firmware-17.168.5.2-38.fc20.noarch
to
iwl6000g2b5-firmware-17.168.5.2-38.fc20.noarch
/usr/lib/firmware/iwlwifi-6000g2b-5.ucode

2. Create
iwl6000g2b6-firmware-18.168.6.1-1.f20.noarch.rpm
/usr/lib/firmware/iwlwifi-6000g2b-6.ucode

- Intel® Centrino® Advanced-N 6205

3. Rename
iwl6000g2a-firmware-17.168.5.3-38.fc20.noarch
to
iwl6000g2a5-firmware-17.168.5.3-38.fc20.noarch
/usr/lib/firmware/iwlwifi-6000g2a-5.ucode

4. Create
iwl6000g2a6-firmware-18.168.6.1-1.f20.noarch.rpm
/usr/lib/firmware/iwlwifi-6000g2a-6.ucode

poma

[Richard Shaw]

Thanks for tips!

I did a yum --enablerepo=updates-testing update kernel* iwl* and got the
latest packages.

The connection definitely seems "better", it didn't take nearly as long to
remote in over ssh but I'm still not seeing the 108Mbit, does that require
a 5GHz connection instead of 2.4GHz?

Thanks,
Richard

[poma]

On 11.06.2014 15:33, Richard Shaw wrote:
> Thanks for tips!
>
> I did a yum --enablerepo=updates-testing update kernel* iwl* and got the
> latest packages.
>
> The connection definitely seems "better", it didn't take nearly as long to
> remote in over ssh but I'm still not seeing the 108Mbit, does that require
> a 5GHz connection instead of 2.4GHz?
>
> Thanks,
> Richard

You can play with iwlwifi parameter '11n_disable'.
However what HW & FW your AP/router is, and what are the settings?

poma

[Richard Shaw]

On Wed, Jun 11, 2014 at 2:01 PM, poma  wrote:
> On 11.06.2014 15:33, Richard Shaw wrote:
> > Thanks for tips!
> >
> > I did a yum --enablerepo=updates-testing update kernel* iwl* and got the
> > latest packages.
> >
> > The connection definitely seems "better", it didn't take nearly as long to
> > remote in over ssh but I'm still not seeing the 108Mbit, does that require
> > a 5GHz connection instead of 2.4GHz?
> >
> > Thanks,
> > Richard
>
> You can play with iwlwifi parameter '11n_disable'.
> However what HW & FW your AP/router is, and what are the settings?

Were is that set? I didn't see it in iwconfig...

The access point is a Ubiquity UniFi AP. The channel is set to auto and
using the Unifi software I can see other devices are connecting as "n"
instead of "g" like the laptop.

Thanks,
Richard

[poma]

On 11.06.2014 21:39, Richard Shaw wrote:
> On Wed, Jun 11, 2014 at 2:01 PM, poma  wrote:
> > On 11.06.2014 15:33, Richard Shaw wrote:
> > > Thanks for tips!
> > >
> > > I did a yum --enablerepo=updates-testing update kernel* iwl* and got the
> > > latest packages.
> > >
> > > The connection definitely seems "better", it didn't take nearly as long to
> > > remote in over ssh but I'm still not seeing the 108Mbit, does that require
> > > a 5GHz connection instead of 2.4GHz?
> > >
> > > Thanks,
> > > Richard
> >
> > You can play with iwlwifi parameter '11n_disable'.
> > However what HW & FW your AP/router is, and what are the settings?
>
> Were is that set? I didn't see it in iwconfig...

:)
$ modinfo -p iwlwifi | grep 11n

- disable full
# modprobe -vr iwlwifi
# modprobe -v iwlwifi 11n_disable=1
# cat /sys/module/iwlwifi/parameters/11n_disable
1

- disable agg TX
# modprobe -vr iwlwifi
# modprobe -v iwlwifi 11n_disable=2
# cat /sys/module/iwlwifi/parameters/11n_disable
2

- disable agg RX
# modprobe -vr iwlwifi
# modprobe -v iwlwifi 11n_disable=4
# cat /sys/module/iwlwifi/parameters/11n_disable
4

- enable agg TX
# modprobe -vr iwlwifi
# modprobe -v iwlwifi 11n_disable=8
# cat /sys/module/iwlwifi/parameters/11n_disable
8
> The access point is a Ubiquity UniFi AP. The channel is set to auto and
> using the Unifi software I can see other devices are connecting as "n"
> instead of "g" like the laptop.
>
> Thanks,
> Richard

There are four "indoor" models, and basic one ain't 5 GHz.
Besides there is no soft for the linux distros.

poma

[Richard Shaw]

On Wed, Jun 11, 2014 at 3:31 PM, poma  wrote:
> There are four "indoor" models, and basic one ain't 5 GHz.

Yes, I have the basic one, so it does support "n" but in 2.4GHz only.
> Besides there is no soft for the linux distros.

The discovery software is java based and does run, but I couldn't get it to
work.

The full unifi software is java with a mongodb database backend and works
fine. I have a RPM I created, the only problem I haven't been able to fix
is the selinux issues, one for the private mongodb instance, and then the
ports it binds to.

Richard

[poma]

On 11.06.2014 22:48, Richard Shaw wrote:
> On Wed, Jun 11, 2014 at 3:31 PM, poma  wrote:
> > There are four "indoor" models, and basic one ain't 5 GHz.
>
> Yes, I have the basic one, so it does support "n" but in 2.4GHz only.
> > Besides there is no soft for the linux distros.
>
> The discovery software is java based and does run, but I couldn't get it to
> work.
>
> The full unifi software is java with a mongodb database backend and works
> fine. I have a RPM I created, the only problem I haven't been able to fix
> is the selinux issues, one for the private mongodb instance, and then the
> ports it binds to.
>
> Richard

OK. there is soft[1] although it is not visible on the official download site, however not officially supported and not fully functional.

$ 7z x -so UniFi.unix.zip UniFi/readme.txt | cat
...
Processing archive: UniFi.unix.zip

Extracting  UniFi/readme.txt

Everything is Ok
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This is UniFi build for Linux/Unix-like system. It is not officially
supported / regularly tested.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

It's basically a package that strips all of the native components

Limitations
===================
* Auto update (of UniFi controller) does not work
* Backup/Restore does not work (see FAQ)

System Requirements
===================
* mongoDB 1.6.5 (http://www.mongodb.org/downloads)
* Sun Java 6

Install
===================
1. unzip it to a directory you prefer, e.g. /path/to/UniFi
2. make sure you have mongod installed (out database server) for your platform
3. make sure /path/to/UniFi/bin/mongod points to the /path/to/mongod
4. to run it,
    cd /path/to/UniFi/
    java -jar lib/ace.jar start
5. to stop it,
    cd /path/to/UniFi/
    java -jar lib/ace.jar stop

...

$ java -jar UniFi/lib/ace.jar start
Port 8443 not available

BTW it seems that the dd-wrt can be installed on it but not for free.
ftp://ftp.dd-wrt.com/betas/2014/05-27-2014-r24160/ubnt_UniFi_AP/install.txt
Besides you do this at your own risk and expense.

And what about the "modprobe iwlwifi 11n_disable=[1248]",
does it have any effect on throughput?

poma

[1] http://goo.gl/lWuQON

[Daniel J Walsh]

On 06/11/2014 01:48 PM, Richard Shaw wrote:
> On Wed, Jun 11, 2014 at 3:31 PM, poma <[e-mail]
> > wrote:
> There are four "indoor" models, and basic one ain't 5 GHz.
>
>
> Yes, I have the basic one, so it does support "n" but in 2.4GHz only.
> Besides there is no soft for the linux distros.
>
>
> The discovery software is java based and does run, but I couldn't get
> it to work.
>
> The full unifi software is java with a mongodb database backend and
> works fine. I have a RPM I created, the only problem I haven't been
> able to fix is the selinux issues, one for the private mongodb
> instance, and then the ports it binds to.
>
> Richard

Please open a bugzilla for the SELinux issues.

[Richard Shaw]

On Thu, Jun 12, 2014 at 6:56 AM, Daniel J Walsh  wrote:
> The full unifi software is java with a mongodb database backend and works
> fine. I have a RPM I created, the only problem I haven't been able to fix
> is the selinux issues, one for the private mongodb instance, and then the
> ports it binds to.
>
> Richard
>
> Please open a bugzilla for the SELinux issues.

Dan,

I know this is something we discussed a bit in the past but I ran out of
steam trying to get it right. Perhaps a BZ ticket would be the better way
of tracking it, thanks.

Richard

[Richard Shaw]

On Thu, Jun 12, 2014 at 6:56 AM, Daniel J Walsh  wrote:
> The full unifi software is java with a mongodb database backend and works
> fine. I have a RPM I created, the only problem I haven't been able to fix
> is the selinux issues, one for the private mongodb instance, and then the
> ports it binds to.
>
> Please open a bugzilla for the SELinux issues.

Before I open a BZ, here's what I have in my spec file which from what I
understand should be persistent...

%posttrans
/usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/logs(/.*)?"
/usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data(/.*)?"
/usr/sbin/semanage port -m -t mongod_port_t 27117

Or should this be handled in a policy?

Thanks,
Richard

[Richard Shaw]

> Install
> unzip it to a directory you prefer, e.g. /path/to/UniFi
> make sure you have mongod installed (out database server) for your
>
> platform
> 3. make sure /path/to/UniFi/bin/mongod points to the /path/to/mongod
> 4. to run it,
>         cd /path/to/UniFi/
>         java -jar lib/ace.jar start
> 5. to stop it,
>         cd /path/to/UniFi/
>         java -jar lib/ace.jar stop

It seems to work OK as long as you have the right context on the data and
log directories.
> $ java -jar UniFi/lib/ace.jar start
> Port 8443 not available

semanage port -m -t mongod_port_t 27117

BTW it seems that the dd-wrt can be installed on it but not for free.
> ftp://ftp.dd-wrt.com/betas/2014/05-27-2014-r24160/ubnt_
> UniFi_AP/install.txt
> Besides you do this at your own risk and expense.
>
> And what about the "modprobe iwlwifi 11n_disable=[1248]",
> does it have any effect on throughput?

In every case it ramped up to 54MBit eventually although it some cases it
was very slow to do so.

Thanks,
Richard

[poma]

On 13.06.2014 04:09, Richard Shaw wrote:
...
> > And what about the "modprobe iwlwifi 11n_disable=[1248]",
> > does it have any effect on throughput?
>
> In every case it ramped up to 54MBit eventually although it some cases it
> was very slow to do so.
>
> Thanks,
> Richard

e.g.
# modprobe -v iwlwifi swcrypto=1
# cat /sys/module/iwlwifi/parameters/swcrypto
1

At least you can attempt to read and apply what's recommended on the router/AP side:
-"How do I improve my 802.11n wireless performance?"
http://www.intel.com/support/wireless/wlan/4965agn/sb/CS-028449.htm
- Wireless-N Configuration
http://www.dd-wrt.com/wiki/index.php/Wireless-N_Configuration

Besides you can always seek advice at http://wireless.kernel.org/en/users/Drivers/iwlwifi#Support

poma

[poma]

Also "Disable 40MHz support in the 2.4GHz band" solo or combined,
e.g.
# modprobe -rv iwlwifi
# modprobe -v cfg80211 cfg80211_disable_40mhz_24ghz=1
# modprobe -v iwlwifi swcrypto=1
# cat /sys/module/cfg80211/parameters/cfg80211_disable_40mhz_24ghz
Y
# cat /sys/module/iwlwifi/parameters/swcrypto
1

poma

[Richard Shaw]

Well thanks for all the pointers and ideas but I think I've got it "fixed"
now... I'm still not sure what caused the breakage though...

I noticed there was a new release of the UniFi software (3.2.1) and managed
to upgrade it and then updated the firmware on the AP, now the laptop is
connecting with 802.11n again.

Thanks,
Richard

[poma]

On 14.06.2014 05:21, Richard Shaw wrote:
> Well thanks for all the pointers and ideas but I think I've got it "fixed"
> now... I'm still not sure what caused the breakage though...
>
> I noticed there was a new release of the UniFi software (3.2.1) and managed
> to upgrade it and then updated the firmware on the AP, now the laptop is
> connecting with 802.11n again.
>
> Thanks,
> Richard

Super duper!
What are the actual speeds achieved with Intel Centrino Advanced-N 6235 i.e. iwlwifi in N-mode?
Min&Max, Up&Down.

poma

[Richard Shaw]

On Sat, Jun 14, 2014 at 7:17 AM, poma  wrote:
> On 14.06.2014 05:21, Richard Shaw wrote:
> > Well thanks for all the pointers and ideas but I think I've got it "fixed"
> > now... I'm still not sure what caused the breakage though...
> >
> > I noticed there was a new release of the UniFi software (3.2.1) and
> > managed
> > to upgrade it and then updated the firmware on the AP, now the laptop is
> > connecting with 802.11n again.
> Super duper!
> What are the actual speeds achieved with Intel Centrino Advanced-N 6235
> i.e. iwlwifi in N-mode?
> Min&Max, Up&Down.

I downloaded a recording from my MythTV box, so plain http file transfer,
and it peaked over 5MB/s! Looking at iwconfig the max connection speed (I
assume this is aggregate) was 144.4Mbit.

I then did an scp transfer from the laptop to my desktop and it averaged
around 2.5MB/s so that's still pretty respectable for 2.4GHz wifi.

The few limitations of the software notwithstanding, Ubiquity seems to make
decent stuff for almost consumer grade prices. I need to do some testing on
a virgin install since I don't want to mess up my working installation but
I plan on packaging the UniFi software for Fedora. Unfortunately due to it
being non-FOSS it will have to go into RPM Fusion non-free but that's
better than not being packaged at all!

Thanks,
Richard

[poma]

On 14.06.2014 15:52, Richard Shaw wrote:
> On Sat, Jun 14, 2014 at 7:17 AM, poma  wrote:
> > On 14.06.2014 05:21, Richard Shaw wrote:
> > > Well thanks for all the pointers and ideas but I think I've got it "fixed"
> > > now... I'm still not sure what caused the breakage though...
> > >
> > > I noticed there was a new release of the UniFi software (3.2.1) and
> > > managed
> > > to upgrade it and then updated the firmware on the AP, now the laptop is
> > > connecting with 802.11n again.
> > Super duper!
> > What are the actual speeds achieved with Intel Centrino Advanced-N 6235
> > i.e. iwlwifi in N-mode?
> > Min&Max, Up&Down.
>
> I downloaded a recording from my MythTV box, so plain http file transfer,
> and it peaked over 5MB/s! Looking at iwconfig the max connection speed (I
> assume this is aggregate) was 144.4Mbit.

Data Rate for MCS index 15 in HT20 mode[1].
> I then did an scp transfer from the laptop to my desktop and it averaged
> around 2.5MB/s so that's still pretty respectable for 2.4GHz wifi.

Actual speed "should" be a "little" faster, considering I've seen higher speeds in G-mode.
However.
> The few limitations of the software notwithstanding, Ubiquity seems to make
> decent stuff for almost consumer grade prices. I need to do some testing on
> a virgin install since I don't want to mess up my working installation but
> I plan on packaging the UniFi software for Fedora. Unfortunately due to it
> being non-FOSS it will have to go into RPM Fusion non-free but that's
> better than not being packaged at all!
>
> Thanks,
> Richard

For the same price category you can get e.g. TL-WDR4310[2] which is BTW covered[4] for free by the dd-wrt(v1.x) and OpenWRT(v1.0), has newer Atheros SoC with additional USB functionality, etc.

poma

[1] http://mcsindex.com/
[2] http://wiki.openwrt.org/toh/tp-link/tl-wdr4310
       http://wiki.openwrt.org/toh/ubiquiti/unifi
[4] http://dd-wrt.com/site/support/router-database

[Bob Goodwin]

On 06/14/14 12:04, poma wrote:
> For the same price category you can get e.g. TL-WDR4310[2] which is 
> BTW covered[4] for free by the dd-wrt(v1.x) and OpenWRT(v1.0), has 
> newer Atheros SoC with additional USB functionality, etc.
>
> poma
>
> [1] http://mcsindex.com/
> [2] http://wiki.openwrt.org/toh/tp-link/tl-wdr4310
>       http://wiki.openwrt.org/toh/ubiquiti/unifi
> [4] http://dd-wrt.com/site/support/router-database

Do you have experience with this [TL-WDR4310] router? I see they claim 
compatibility with DD-WRT, I prefer the Tomato version which I have on 
all my routers, this one is a Cisco E3000.

Bob

[poma]

On 14.06.2014 18:21, Bob Goodwin - Zuni, Virginia, USA wrote:
> On 06/14/14 12:04, poma wrote:
> > For the same price category you can get e.g. TL-WDR4310[2] which is
> > BTW covered[4] for free by the dd-wrt(v1.x) and OpenWRT(v1.0), has
> > newer Atheros SoC with additional USB functionality, etc.
> >
> > poma
> >
> > [1] http://mcsindex.com/
> > [2] http://wiki.openwrt.org/toh/tp-link/tl-wdr4310
> >        http://wiki.openwrt.org/toh/ubiquiti/unifi
> > [4] http://dd-wrt.com/site/support/router-database
>
> Do you have experience with this [TL-WDR4310] router? I see they claim
> compatibility with DD-WRT, I prefer the Tomato version which I have on
> all my routers, this one is a Cisco E3000.
>
> Bob

Nope for this particular model, it's just one example for comparison.
BTW dd-wrt is versatile and covers a lot more devices, besides the excellent documentation.
Your  Linksys E3000 is literally twice as expensive, if a new one can be purchased, at all.
It's not for everyone's pocket, Bob!

poma

[Bob Goodwin]

On 06/14/14 13:02, poma wrote:
> Nope for this particular model, it's just one example for comparison.
> BTW dd-wrt is versatile and covers a lot more devices, besides the 
> excellent documentation.
> Your  Linksys E3000 is literally twice as expensive, if a new one can 
> be purchased, at all.
> It's not for everyone's pocket, Bob!
>
> poma

I just thought you might have known of an exceptional buy. I always look 
at router spec's in the Newegg ads.

The E3000 I am using was a "refurbished" one I bought perhaps two years 
ago and has never missed a beat. I had another I paid full price for and 
somehow "bricked" it, the remains lay at the bottom of my junk box, I 
really regretted that!

The Tomato (beta) Version 1.28 software provides information about 
signal quality, etc. and gives me a lot of control over usage which is a 
necessity with my Viasat service.

Thanks,

Bob

[poma]

On 14.06.2014 19:19, Bob Goodwin - Zuni, Virginia, USA wrote:
> On 06/14/14 13:02, poma wrote:
> > Nope for this particular model, it's just one example for comparison.
> > BTW dd-wrt is versatile and covers a lot more devices, besides the
> > excellent documentation.
> > Your  Linksys E3000 is literally twice as expensive, if a new one can
> > be purchased, at all.
> > It's not for everyone's pocket, Bob!
> >
> > poma
>
> I just thought you might have known of an exceptional buy. I always look
> at router spec's in the Newegg ads.
>
> The E3000 I am using was a "refurbished" one I bought perhaps two years
> ago and has never missed a beat. I had another I paid full price for and
> somehow "bricked" it, the remains lay at the bottom of my junk box, I
> really regretted that!

Are you really sure it is "bricked".
If to you brand and design are not so important, see for yourself on egg how much is new TL-WDR4300 and on linksys store how much is, not a new one, but refurbished E3000. :)
> The Tomato (beta) Version 1.28 software provides information about
> signal quality, etc. and gives me a lot of control over usage which is a
> necessity with my Viasat service.
>
> Thanks,
>
> Bob

If Tomato suits you well, let it be! :)

poma

[poma]

On 14.06.2014 20:31, poma wrote:
> On 14.06.2014 19:19, Bob Goodwin - Zuni, Virginia, USA wrote:
> > On 06/14/14 13:02, poma wrote:
> > > Nope for this particular model, it's just one example for comparison.
> > > BTW dd-wrt is versatile and covers a lot more devices, besides the
> > > excellent documentation.
> > > Your  Linksys E3000 is literally twice as expensive, if a new one can
> > > be purchased, at all.
> > > It's not for everyone's pocket, Bob!
> > >
> > > poma
> >
> > I just thought you might have known of an exceptional buy. I always look
> > at router spec's in the Newegg ads.
> >
> > The E3000 I am using was a "refurbished" one I bought perhaps two years
> > ago and has never missed a beat. I had another I paid full price for and
> > somehow "bricked" it, the remains lay at the bottom of my junk box, I
> > really regretted that!
>
> Are you really sure it is "bricked".
> If to you brand and design are not so important, see for yourself on egg how much is new TL-WDR4300 and on linksys store how much is, not a new one, but refurbished E3000. :)
> > The Tomato (beta) Version 1.28 software provides information about
> > signal quality, etc. and gives me a lot of control over usage which is a
> > necessity with my Viasat service.
> >
> > Thanks,
> >
> > Bob
>
> If Tomato suits you well, let it be! :)

However, since you are a fan of Tomato and it is dedicated for routers with Broadcom SoC, the Black Diamond series RT-N66U is probably a good choice for you.

Tomato by Shibby
http://tomato.groov.pl

poma

[Daniel J Walsh]

On 06/12/2014 10:14 AM, Richard Shaw wrote:
> On Thu, Jun 12, 2014 at 6:56 AM, Daniel J Walsh <[e-mail]
> > wrote:
> > The full unifi software is java with a mongodb database backend
> > and works fine. I have a RPM I created, the only problem I
> > haven't been able to fix is the selinux issues, one for the
> > private mongodb instance, and then the ports it binds to. 
> Please open a bugzilla for the SELinux issues.
>
>
> Before I open a BZ, here's what I have in my spec file which from what
> I understand should be persistent...
>
> %posttrans
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/logs(/.*)?"
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data(/.*)?"
> /usr/sbin/semanage port -m -t mongod_port_t 27117
>
> Or should this be handled in a policy?
>
> Thanks,
> Richard

I think your post install should look like.

/usr/sbin/semanage fcontext -e /var/log/mongod "/var/lib/unifi/logs"
/usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data"
/usr/sbin/semanage port -m -t mongod_port_t 27117

Don't use the regex. Also I would figure the logs should be labeled
mongod_log_t rather then mongod_lib_t.

If this is a standard location for this code, we should put it into the
base package.

[Richard Shaw]

On Mon, Jun 16, 2014 at 12:19 PM, Daniel J Walsh  wrote:
> On 06/12/2014 10:14 AM, Richard Shaw wrote:
>
> On Thu, Jun 12, 2014 at 6:56 AM, Daniel J Walsh 
> wrote:
> > The full unifi software is java with a mongodb database backend and
> >
> > works fine. I have a RPM I created, the only problem I haven't been able to
> > fix is the selinux issues, one for the private mongodb instance, and then
> > the ports it binds to.
> >
> > Please open a bugzilla for the SELinux issues.
>
> Before I open a BZ, here's what I have in my spec file which from what I
> understand should be persistent...
>
> %posttrans
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/logs(/.*)?"
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data(/.*)?"
> /usr/sbin/semanage port -m -t mongod_port_t 27117
>
> Or should this be handled in a policy?
>
> Thanks,
> Richard
>
> I think your post install should look like.
>
> /usr/sbin/semanage fcontext -e /var/log/mongod "/var/lib/unifi/logs"
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data"
> /usr/sbin/semanage port -m -t mongod_port_t 27117
>
> Don't use the regex. Also I would figure the logs should be labeled
> mongod_log_t rather then mongod_lib_t.

What is the concern with regex? It is specific to packaging? Most of the
examples I found online used that method... As far as the label, since
everything is getting dumped in /var/lib I figured that would be OK.

If this is a standard location for this code, we should put it into the
> base package.

There is not a standard install location, the install will "work" as long
as everything stays in the same relative location (the unifi directory).
Since it writes a lot of stuff I figured /var was the best (only?) real
option.

Following the example of a draft wiki I can't find anymore I had modified
the scripts to this instead of using %posttrans:
%post
semanage fcontext -a -t mongod_var_lib_t \
    "%{_sharedstatedir}/unifi/logs(/.*)?" 2>/dev/null || :
semanage fcontext -a -t mongod_var_lib_t \
    "%{_sharedstatedir}/unifi/data(/.*)?" 2>/dev/null || :
restorecon -R %{_sharedstatedir}/unifi/logs || :
restorecon -R %{_sharedstatedir}/unifi/data || :
semanage port -m -t mongod_port_t 27117 || :

%postun
if [ $1 -eq 0 ] ; then  # final removal
semanage fcontext -d -t mongod_var_lib_t \
    "%{_sharedstatedir}/unifi/logs(/.*)?" 2>/dev/null || :
semanage fcontext -d -t mongod_var_lib_t \
    "%{_sharedstatedir}/unifi/data(/.*)?" 2>/dev/null || :
fi

Thanks,
Richard

[Daniel J Walsh]

On 06/16/2014 01:35 PM, Richard Shaw wrote:
> On Mon, Jun 16, 2014 at 12:19 PM, Daniel J Walsh <[e-mail]
> > wrote:
> On 06/12/2014 10:14 AM, Richard Shaw wrote:
> > On Thu, Jun 12, 2014 at 6:56 AM, Daniel J Walsh
> > <[e-mail] <mailto:[e-mail]>> wrote:
> >
> > >     The full unifi software is java with a mongodb database
> > >     backend and works fine. I have a RPM I created, the only
> > >     problem I haven't been able to fix is the selinux issues,
> > >     one for the private mongodb instance, and then the ports it
> > >     binds to. 
> >     Please open a bugzilla for the SELinux issues.
> >
> >
> > Before I open a BZ, here's what I have in my spec file which from
> > what I understand should be persistent...
> >
> > %posttrans
> > /usr/sbin/semanage fcontext -e /var/lib/mongod
> > "/var/lib/unifi/logs(/.*)?"
> > /usr/sbin/semanage fcontext -e /var/lib/mongod
> > "/var/lib/unifi/data(/.*)?"
> > /usr/sbin/semanage port -m -t mongod_port_t 27117
> >
> > Or should this be handled in a policy?
> >
> > Thanks,
> > Richard
> I think your post install should look like.
>
> /usr/sbin/semanage fcontext -e /var/log/mongod "/var/lib/unifi/logs"
> /usr/sbin/semanage fcontext -e /var/lib/mongod "/var/lib/unifi/data"
> /usr/sbin/semanage port -m -t mongod_port_t 27117
>
> Don't use the regex. Also I would figure the logs should be
> labeled mongod_log_t rather then mongod_lib_t.
>
>
> What is the concern with regex?
> It is specific to packaging? Most of the examples I found online used
> that method... As far as the label, since everything is getting dumped
> in /var/lib I figured that would be OK.

Not a concern with regex. it just will not work.  The examples you have
seen on line, were not using equivalence.  They were using generic
labelling.

Equivalence tells SELinux to swap the second part of the path with the
first.  You code would only match file paths that began with
/var/lib/unifi/logs(/.*?)  Not /var/lib/unifi/logs/foobar.log
> If this is a standard location for this code, we should put it
> into the base package.
>
>
> There is not a standard install location, the install will "work" as
> long as everything stays in the same relative location (the unifi
> directory). Since it writes a lot of stuff I figured /var was the best
> (only?) real option.

Yes
> Following the example of a draft wiki I can't find anymore I had
> modified the scripts to this instead of using %posttrans:
> %post
> semanage fcontext -a -t mongod_var_lib_t \
>     "%{_sharedstatedir}/unifi/logs(/.*)?" 2>/dev/null || :
> semanage fcontext -a -t mongod_var_lib_t \
>     "%{_sharedstatedir}/unifi/data(/.*)?" 2>/dev/null || :
> restorecon -R %{_sharedstatedir}/unifi/logs || :
> restorecon -R %{_sharedstatedir}/unifi/data || :
> semanage port -m -t mongod_port_t 27117 || :
>
> %postun
> if [ $1 -eq 0 ] ; then  # final removal
> semanage fcontext -d -t mongod_var_lib_t \
>     "%{_sharedstatedir}/unifi/logs(/.*)?" 2>/dev/null || :
> semanage fcontext -d -t mongod_var_lib_t \
>     "%{_sharedstatedir}/unifi/data(/.*)?" 2>/dev/null || :
> fi
>
> Thanks,
> Richard

That should work.  You could speed it up by combining both semange
fcontext lines into a single transaction. Something like.

semanage -S targeted -i - << _EOF
fcontext -a -t mongod_var_lib_t "%{_sharedstatedir}/unifi/logs(/.*)?"
fcontext -a -t mongod_var_lib_t "%{_sharedstatedir}/unifi/data(/.*)?"
_EOF 2>/dev/null || :