[j-nsp] Sandisk Extreme III 1G CF card compatibility

Dave Diller dave at maxgigapop.net
Mon May 5 13:14:41 EDT 2008


This is an FYI, and a separate issue form Pekka's discovery that these  
faster cards can hog CPU when being written.

It turns out that some SanDisk Extreme III 1G cards are Juniper  
friendly, and others are not.  We purchased three at the same time  
from the same CompUSA.  Of the three, two do not work and the third  
one is fine. The symptom is that the non-functional ones time out and  
you can't install anything on them (but 6509's are fine with them, so  
they'll be redeployed there ;-):

=================== Bootstrap installer starting ===================
Initialized the environment
Routing engine model is RE-4.0
Discovered that flash disk = ad0 , hard disk = ad1
ad0: TIMEOUT - READ_MUL retrying (2 retries left) LBA=79
ad0: TIMEOUT - READ_MUL retrying (1 retry left) LBA=79
ad0: TIMEOUT - READ_MUL retrying (0 retries left) LBA=79

WARNING: The installation will erase the contents of your disks.
Do you wish to continue (y/n)?
y
Setting ospackage=jboot-9.0R2.10.tgz, configpackage=
Setting packlist=jbundle-9.0R2.10-domestic.tgz
Packages mounted from the install media
Partitioning ad0...
******* Working on device /dev/ad0 *******
Installing disk label on ad0
Running newfs on ad0s1a...
/dev/ad0s1a: 879.7MB (1801640 sectors) block size 16384, fragment size  
2048
	using 5 cylinder groups of 183.69MB, 11756 blks, 23552 inodes.
super-block backups (for fsck -b #) at:
32, 376224, 752416, 1128608, 1504800
ad0: TIMEOUT - READ_MUL retrying (2 retries left) LBA=127
ad0: TIMEOUT - READ_MUL retrying (1 retry left) LBA=127
ad0: TIMEOUT - READ_MUL retrying (0 retries left) LBA=127
*snip*
and it keeps trying and timing out for 20 minutes or so.



This one:
ad0: 977MB <SanDisk SDCFX-1024 HDX 4.03> at ata0-master PIO4
is the one that works.  The card has ID BB061004765B printed on the  
edge.

The two that do not have code BB070504765B, and they ID on boot as
ad0: 977MB <SanDisk SDCFX3-1024 HDX 4.08> at ata0-master PIO4

The difference in the beginning portion of the printed code looks like  
a date code.  The 4.03 version looks like late 2006, and the 4.08  
early 2007.  So old stock seems fine; later ones do not.

FWIW, the older one that works has "ESP technology" printed UNDER the  
1.0GB part of the label, while the newer ones have the ESP logo to the  
LEFT of the 1.0GB.

-dd



More information about the juniper-nsp mailing list