Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-07 Thread Seth Mos

Op 6-3-2011 23:26, Bao Ha schreef:

Hi Bart,

Thanks for the note.

According to the forum, it should not be a problem. :-(


It is not.


When we first got the reports of corrupted CFs, we just overnighted new
ones. Then, those died shortly, within a week or two. We replaced a
complete system: systemboard, memory and CF.


Why are you shipping cheap CF cards without wear levelling?

I have run a full install on a Lexar 1GB CF for over 4 years before the 
CF card died.


I've also run into the CF without wear levelling issue. Get a proper 
CF card.


Regards,
Seth

-
To unsubscribe, e-mail: support-unsubscr...@pfsense.com
For additional commands, e-mail: support-h...@pfsense.com

Commercial support available - https://portal.pfsense.org



Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-07 Thread Bao Ha
Hi Seth,

On Mon, Mar 7, 2011 at 12:05 AM, Seth Mos seth@dds.nl wrote:

 Op 6-3-2011 23:26, Bao Ha schreef:

  Hi Bart,

 Thanks for the note.

 According to the forum, it should not be a problem. :-(


 It is not.


I have had three systems with corrupted flash memory: two with the Kingston
4GB Elite Pro, one with a 4GB flash drive.



  When we first got the reports of corrupted CFs, we just overnighted new
 ones. Then, those died shortly, within a week or two. We replaced a
 complete system: systemboard, memory and CF.


 Why are you shipping cheap CF cards without wear levelling?


We used to offer a choice of CF or DOM.The DOM has industrial-strength
wear-leveling. It was also better since in the early days, our systemboards
choked on DMA with faster CF cards. Nobody wants DOM!

I did not rule out that Kingston's quality may have dropped significantly.
If that is the case, I'll switch to a different brand name.

I have run a full install on a Lexar 1GB CF for over 4 years before the CF
 card died.

 I've also run into the CF without wear levelling issue. Get a proper CF
 card.


We have been shipping more than a thousand systems with Kingston CF since
2006 with no corrupted flash memory.

What ever killed the two Kingston Elite pro 4GB CFs within two weeks will
also kill a DOM or industrial CF, maybe not in weeks or months, but probably
within a year.

Bao



 Regards,
 Seth


 -
 To unsubscribe, e-mail: support-unsubscr...@pfsense.com
 For additional commands, e-mail: support-h...@pfsense.com

 Commercial support available - https://portal.pfsense.org




-- 
Best Regards.
Bao C. Ha
Hacom - Embedded Systems and Appliances
http://www.hacom.net
voice: (714) 564-9932


Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-07 Thread Vick Khera
On Sun, Mar 6, 2011 at 5:05 PM, Bao Ha b...@hacom.net wrote:
 Something happened in BETA5 and it was carried into RC1, up to today
 snapshot: 20110306-0859.


I see this in my embedded BETA5 install at home (I should upgrade soon
to RC1 I suppose...)

I see no significant amount of writing to it.  There are no extra
packages installed and all it does is basic NAT + firewall + IPsec
VPN.  It is a fairly generic CF card too.

-
To unsubscribe, e-mail: support-unsubscr...@pfsense.com
For additional commands, e-mail: support-h...@pfsense.com

Commercial support available - https://portal.pfsense.org



RE: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-06 Thread Bart Grefte
Someone already made a bugreport  http://redmine.pfsense.org/issues/1279
http://redmine.pfsense.org/issues/1279  ;)

 

 

  _  

Van: Bao Ha [mailto:b...@hacom.net] 
Verzonden: zondag 6 maart 2011 23:06
Aan: customersupp...@pfsense.org
CC: support@pfsense.com
Onderwerp: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

 

Something happened in BETA5 and it was carried into RC1, up to today
snapshot: 20110306-0859.


The file system in nanobsd version is now mounted fully RW, see the
following mount command:

...

[2.0-RC1][admin@pfHacom.localdomain]/root(1): mount

/dev/ufs/pfsense0 on / (ufs, local)


devfs on /dev (devfs, local)


/dev/md0 on /tmp (ufs, local)


/dev/md1 on /var (ufs, local)


/dev/ufs/cf on /cf (ufs, local)


devfs on /var/dhcpd/dev (devfs, local)

...

 

I believe they are supposed to be mounted read-only or at least RW with
NOATIME.

 

We have had at least two systems running pfSense 2.0 BETA5 and RC1 RMAed
back with suspected hardware problems, causing corruption of compact flash
memory. We think the root cause of this problem is due to the filesystems
mounted fully RW in the compact flash.

 

We plan to distribute the following temporary fix to our custmers who want
to run pfSense 2.0:

...

[2.0-RC1][admin@pfHacom.localdomain]/root(1): cat
/usr/local/etc/rc.d/hacom.sh

#!/bin/sh


 


# hacom.sh - BCH 3/6/2011


#   Temprorary fix to mount the filesystem Read-Only to avoid destroying
flash memory 

 


PLATFORM=`/bin/cat /etc/platform`


 


if [ $PLATFORM = nanobsd ]; then


/sbin/mount -u -oro /; /sbin/mount -u -onoatime /cf


fi


...

 

Appreciate if someone look into this problem.

 

I have also CCed this message to support@pfsense.com to notify others
currently using pfSense 2.0 RC1 nanobsd version of the danger to flash
memory.

 

Thanks.

Bao

-- 
Best Regards.
Bao C. Ha
Hacom - Embedded Systems and Appliances
http://www.hacom.net 
voice: (714) 564-9932



Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-06 Thread Bao Ha
Hi Bart,

Thanks for the note.

According to the forum, it should not be a problem. :-(

Unfortunately, mounting RW without NOATIME will pounce on the compact flash
everytime a READ is made. It will kill the CF sooner or later.

When we first got the reports of corrupted CFs, we just overnighted new
ones. Then, those died shortly, within a week or two. We replaced a complete
system: systemboard, memory and CF.

I am loosing my hair and sleeps, thinking one of our most reliable systems
being shipped since 2006 is having compatibility issues with pfSense 2.0.

I am hoping that this is the real cause. And I can stop a flood of support
issues.

Bao

On Sun, Mar 6, 2011 at 2:12 PM, Bart Grefte b...@ravenslair.nl wrote:

  Someone already made a bugreport http://redmine.pfsense.org/issues/1279
  ;)




  --

 *Van:* Bao Ha [mailto:b...@hacom.net]
 *Verzonden:* zondag 6 maart 2011 23:06
 *Aan:* customersupp...@pfsense.org
 *CC:* support@pfsense.com
 *Onderwerp:* [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem



 Something happened in BETA5 and it was carried into RC1, up to today
 snapshot: 20110306-0859.

 The file system in nanobsd version is now mounted fully RW, see the
 following mount command:

 ...

 [2.0-RC1][admin@pfHacom.localdomain]/root(1): mount

 /dev/ufs/pfsense0 on / (ufs, local)


 devfs on /dev (devfs, local)


 /dev/md0 on /tmp (ufs, local)


 /dev/md1 on /var (ufs, local)


 /dev/ufs/cf on /cf (ufs, local)


 devfs on /var/dhcpd/dev (devfs, local)

 ...



 I believe they are supposed to be mounted read-only or at least RW with
 NOATIME.



 We have had at least two systems running pfSense 2.0 BETA5 and RC1 RMAed
 back with suspected hardware problems, causing corruption of compact flash
 memory. We think the root cause of this problem is due to the filesystems
 mounted fully RW in the compact flash.



 We plan to distribute the following temporary fix to our custmers who want
 to run pfSense 2.0:

 ...

 [2.0-RC1][admin@pfHacom.localdomain]/root(1): cat
 /usr/local/etc/rc.d/hacom.sh

 #!/bin/sh





 # hacom.sh - BCH 3/6/2011


 #   Temprorary fix to mount the filesystem Read-Only to avoid destroying
 flash memory




 PLATFORM=`/bin/cat /etc/platform`





 if [ $PLATFORM = nanobsd ]; then


 /sbin/mount -u -oro /; /sbin/mount -u -onoatime /cf


 fi


 ...



 Appreciate if someone look into this problem.



 I have also CCed this message to support@pfsense.com to notify others
 currently using pfSense 2.0 RC1 nanobsd version of the danger to flash
 memory.



 Thanks.

 Bao

 --
 Best Regards.
 Bao C. Ha
 Hacom - Embedded Systems and Appliances
 http://www.hacom.net
 voice: (714) 564-9932




-- 
Best Regards.
Bao C. Ha
Hacom - Embedded Systems and Appliances
http://www.hacom.net
voice: (714) 564-9932


RE: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-06 Thread Bart Grefte
Hi Bao,

 

You're welcome :)

 

I've read that, but not sure if that is actually true in all cases.

 

Wow, that is fast! I doubt pfSense writes so much in that time the CF-cards
start dieing, although I might be wrong.

Could be the (lack of?) quality of the CF-cards combined with that problem
that is causing them to fail so fast. (This is just me thinking out loud.)

 

Out of curiosity, why ship systems with an OS that is still beta? Well, RC1
now, but still

 

Not sure if this will help, but maybe adding /etc/rc.conf_mount_ro to a
script that runs during boot-up will do some good. It's the command to mount
read-only.

I still have to add that one and /etc/rc.conf_mount_rw to the script that
makes an IPv6 tunnel on my pfSense v1.2.3 system, since the script needs to
write something during the boot of pfSense when the script is started but
can't do that because of RO filesystem

 

Hope this problem will be solved soon!

 

With regards,

 

Bart

 

 

  _  

Van: Bao Ha [mailto:b...@hacom.net] 
Verzonden: zondag 6 maart 2011 23:26
Aan: support@pfsense.com
Onderwerp: Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

 

Hi Bart,

 

Thanks for the note.

 

According to the forum, it should not be a problem. :-(

 

Unfortunately, mounting RW without NOATIME will pounce on the compact flash
everytime a READ is made. It will kill the CF sooner or later.

 

When we first got the reports of corrupted CFs, we just overnighted new
ones. Then, those died shortly, within a week or two. We replaced a complete
system: systemboard, memory and CF.

 

I am loosing my hair and sleeps, thinking one of our most reliable systems
being shipped since 2006 is having compatibility issues with pfSense 2.0.

 

I am hoping that this is the real cause. And I can stop a flood of support
issues.

 

Bao

On Sun, Mar 6, 2011 at 2:12 PM, Bart Grefte b...@ravenslair.nl wrote:

Someone already made a bugreport  http://redmine.pfsense.org/issues/1279
http://redmine.pfsense.org/issues/1279  ;)

 

 

  _  

Van: Bao Ha [mailto:b...@hacom.net] 
Verzonden: zondag 6 maart 2011 23:06
Aan: customersupp...@pfsense.org
CC: support@pfsense.com
Onderwerp: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

 

Something happened in BETA5 and it was carried into RC1, up to today
snapshot: 20110306-0859.


The file system in nanobsd version is now mounted fully RW, see the
following mount command:

...

[2.0-RC1][admin@pfHacom.localdomain]/root(1): mount

/dev/ufs/pfsense0 on / (ufs, local)


devfs on /dev (devfs, local)


/dev/md0 on /tmp (ufs, local)


/dev/md1 on /var (ufs, local)


/dev/ufs/cf on /cf (ufs, local)


devfs on /var/dhcpd/dev (devfs, local)

...

 

I believe they are supposed to be mounted read-only or at least RW with
NOATIME.

 

We have had at least two systems running pfSense 2.0 BETA5 and RC1 RMAed
back with suspected hardware problems, causing corruption of compact flash
memory. We think the root cause of this problem is due to the filesystems
mounted fully RW in the compact flash.

 

We plan to distribute the following temporary fix to our custmers who want
to run pfSense 2.0:

...

[2.0-RC1][admin@pfHacom.localdomain]/root(1): cat
/usr/local/etc/rc.d/hacom.sh

#!/bin/sh


 


# hacom.sh - BCH 3/6/2011


#   Temprorary fix to mount the filesystem Read-Only to avoid destroying
flash memory 

 


PLATFORM=`/bin/cat /etc/platform`


 


if [ $PLATFORM = nanobsd ]; then


/sbin/mount -u -oro /; /sbin/mount -u -onoatime /cf


fi


...

 

Appreciate if someone look into this problem.

 

I have also CCed this message to support@pfsense.com to notify others
currently using pfSense 2.0 RC1 nanobsd version of the danger to flash
memory.

 

Thanks.

Bao

-- 
Best Regards.
Bao C. Ha
Hacom - Embedded Systems and Appliances
http://www.hacom.net 
voice: (714) 564-9932




-- 
Best Regards.
Bao C. Ha
Hacom - Embedded Systems and Appliances
http://www.hacom.net 
voice: (714) 564-9932



Re: [pfSense Support] pfSense 2.0 RC1 Nanobsd Problem

2011-03-06 Thread Nenhum_de_Nos

On Sun, March 6, 2011 19:26, Bao Ha wrote:
 Hi Bart,

 Thanks for the note.

 According to the forum, it should not be a problem. :-(

 Unfortunately, mounting RW without NOATIME will pounce on the compact
 flash
 everytime a READ is made. It will kill the CF sooner or later.

 When we first got the reports of corrupted CFs, we just overnighted new
 ones. Then, those died shortly, within a week or two. We replaced a
 complete
 system: systemboard, memory and CF.

 I am loosing my hair and sleeps, thinking one of our most reliable systems
 being shipped since 2006 is having compatibility issues with pfSense 2.0.

 I am hoping that this is the real cause. And I can stop a flood of support
 issues.

I see this in a 4g nano image, but is no problem as I use microdrive. so,
in case when this be corrected, how can I make it this way ?

is the nano image the best for a microdrive soekris ? I want the full pc
install just using serial instead of vga.

thanks,

matheus

-- 
We will call you cygnus,
The God of balance you shall be

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?

http://en.wikipedia.org/wiki/Posting_style

-
To unsubscribe, e-mail: support-unsubscr...@pfsense.com
For additional commands, e-mail: support-h...@pfsense.com

Commercial support available - https://portal.pfsense.org