I think it depends on own you are using the CF card. If you are logging
something for instance, then you may exceed the RW limit of the CF card in
a very short time (typically 1000000 times). Imagine you have 10 access
per second, in 28 hours you CF is dead !
Kochise
Robert Muil wrote:
Kochise,
That was exactly my assumption, but it is strange that it seems to be
reproducible with all CFs, and on cards that previously tested fine. Also,
the cluster of bad blocks is not always in the same location on disk.
Robert.
Only if you are logging to the same part of the flash.
The time for rewriting the whole card would take many many times longer than
28hrs.
Also, these problems have been observed on a CF card used only for the
operating system. No logging was done to it at all.
Robert.
“Kochise” <kochise@caramail.com> wrote in message
news:cnv5fc$ogn$1@inn.qnx.com…
I think it depends on own you are using the CF card. If you are logging
something for instance, then you may exceed the RW limit of the CF card in
a very short time (typically 1000000 times). Imagine you have 10 access
per second, in 28 hours you CF is dead !
Kochise
Robert Muil wrote:
Kochise,
That was exactly my assumption, but it is strange that it seems to be
reproducible with all CFs, and on cards that previously tested fine.
Also,
the cluster of bad blocks is not always in the same location on disk.
Robert.
Mmmmwell, I gave my 2 cents, now what exactly happened, I don’t know 
Ask Fox Mulder 
Kochise
Robert Muil wrote:
Only if you are logging to the same part of the flash.
The time for rewriting the whole card would take many many times longer than
28hrs.
Also, these problems have been observed on a CF card used only for the
operating system. No logging was done to it at all.
Robert.
“Kochise” <> kochise@caramail.com> > wrote in message
news:cnv5fc$ogn$> 1@inn.qnx.com> …
I think it depends on own you are using the CF card. If you are logging
something for instance, then you may exceed the RW limit of the CF card in
a very short time (typically 1000000 times). Imagine you have 10 access
per second, in 28 hours you CF is dead !
Kochise
Robert Muil wrote:
Kochise,
That was exactly my assumption, but it is strange that it seems to be
reproducible with all CFs, and on cards that previously tested fine.
Also,
the cluster of bad blocks is not always in the same location on disk.
Robert.
Anyone know if there is a way to get a report on number of blocks read/written to a block device? Like what netstat -s provides for network traffic.
Kochise wrote:
Mmmmwell, I gave my 2 cents, now what exactly happened, I don’t know
Ask Fox Mulder >
Kochise
Robert Muil wrote:
Only if you are logging to the same part of the flash.
The time for rewriting the whole card would take many many times longer than
28hrs.
Also, these problems have been observed on a CF card used only for the
operating system. No logging was done to it at all.
Evan Hillas <blarg@blarg.blarg> wrote:
EH > Anyone know if there is a way to get a report on number of blocks read/written to a block device? Like what netstat -s provides for network traffic.
In the QNX4 world there was a useful utility called fsysinfo.
But I don’t think it exists for QNX6.
Even if you were willing to write one, I don’t think that there is a defined
API to get that information from a resource manager.
Maybe, I’m wrong.