Re: My CE-507 gives me Linux bash# prompt. Corrupted flash??

beyondautomation · ‎09-21-2004

My CE-507 gives me Linux bash# prompt. Corrupted flash??

How can I reload the software, hopefully ACNS 5, and the flash image?

I found an ACNS CD rescue.ISO image in the download page and I burned a bootable CD and hooked up a SCSI CDROM to the CE-507 trying to boot from here to "rescue". The SCSI utility sees a bootable CD but it outputs no prompt for me to go forward.

Any expert out there has clue, please?

Here is the console output with the mid section truncated.

Thanks for your help in advance.

============================================

Initializing memory. Please wait.

BIOS Version: CE500 01.11

BIOS Build date: 07/27/00

Symbios, Inc. SDMS (TM) V4.0 PCI SCSI BIOS, PCI Rev. 2.0, 2.1

PCI-4.11.00

HBA ID LUN VENDOR PRODUCT REV SYNC WIDE INT13 CYL/ HD/SEC

--- -- --- -------- ---------------- ---- ---- ---- ----- ------------

0 0 0 IBM DDYS-T18350M S96H 80.0 16 BOOT 1024/255/63

0 7 0 Symbios SYM53C895 0002 80.0 16

Symbios, Inc. PCI boot ROM successfully installed!

Cisco CE Booting From Flash.

running robin2_start.S

This is cmain! You are now running C code!

build date: Wed Sep 13 11:15:08 PDT 2000

build number: 2

Directory verified ok!

phase2: examining phase3 boot-rom: 'default_bootrom'

cmain: going to jump to location PHASE3_MEM_LOCATION = 16777216

cmain: 10 code bytes:

85 -119 -27 -125 -20 16 87 86 83 -24

defphase3.c: Running third-phase bootrom code!

defphase3.c: directory is ok!

defphase3.c: db_dir_get_location("rescue") failed: -5

defphase3.c: your devimg does not seem to have a rescue image in it

defphase3.c: running default system image instead

defphase3.c: default system image is 'default_sysimg'

defphase3.c: default_sysimg: 14 sectors

defphase3.c: loaded 14 device sectors to 0x2000000!

defphase3.c: bzimage checksum: 0x2000400, len = 933386

defphase3.c: initrd checksum: 0x20e4400, len = 844896

defphase3.c: kernel checksum = 50a8da8f2ab55fe3229110f168965f96

defphase3.c: initrd checksum = 2a8a4b2bbe7361edf7696f317f4eaa2b

defphase3.c: calling execute_kernel!

execute_kernel: system image header:

....

..

.. TRUNCATED

..

....

3c59x.c:v0.99H 11/17/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html

eepro100.c:v1.09j 7/27/99 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html

eth0: Intel PCI EtherExpress Pro100 at 0x98804000, 00:04:4D:E3:62:09, IRQ 9.

Board assembly 668081-002, Physical connectors present: RJ45

Primary interface chip i82555 PHY #1.

General self-test: passed.

Serial sub-system self-test: passed.

Internal registers self-test: passed.

ROM checksum self-test: passed (0x04f4518b).

Receiver lock-up workaround activated.

eth1: Intel PCI EtherExpress Pro100 at 0x98806000, 00:04:4D:E3:62:0A, IRQ 10.

Board assembly 668081-002, Physical connectors present: RJ45

Primary interface chip i82555 PHY #1.

General self-test: passed.

Serial sub-system self-test: passed.

Internal registers self-test: passed.

ROM checksum self-test: passed (0x04f4518b).

Receiver lock-up workaround activated.

eepro100.c:v1.09j 7/27/99 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html

Partition check:

sda:

Cisco CacheEngine Flash Driver Version 0.01

wccp: v1.00 (20000327), debug=0

wccp_list_bypass v1.0.0

wccp_list_bypass tuple allocated 20480

DIAMOND-RAMDISK: drd_open called w/ dev 0x0

EXT2-fs warning: checktime reached, running e2fsck is recommended

VFS: Mounted root (ext2 filesystem).

Freeing unused kernel memory: 400k freed

bash#

derekc · ‎09-26-2004

Would truly appreciate if any experts out there can shed some light on this issue. I am stucked with a dead CE-507.

Thanks,

didyap · ‎09-28-2004

Here is a document that may be of help.

http://www.cisco.com/warp/public/117/cache_engine/ce_diskrec.html

beyondautomation · ‎09-28-2004

Unfortunately it doesn't even give me the chance to break in and it boots with a totally different unix/linus like OS ended with bash# prompt.

Any other counters the same problem?

twaite · ‎10-08-2004

I, as well, have the same problem and have looked all over Cisco's site as well as the internet. Problem is the documentaion on these is very poor to non-existant. If you do find a solution, please post.

beyondautomation · ‎10-08-2004

Can any Cisco Engineers out there be kind enough to give us a clue of how to repair this kind of common problem??

twaite · ‎10-08-2004

Just spoke with tech support, and according to the tech I spoke with, if we are booting it the the bash# that means our content engine flash is corrupted and the devices are unrepairable. It seems as if Cisco themselves have no clue to this "emerging technology" as the support is surley lacking on these devices. I would double check yourself, but if this is in fact true then we have useless CDNs.

beyondautomation · ‎10-09-2004

Cisco will need to be responsible for this software design/implementation problem. Cisco will need to provide a solution (software re-installation procedures or hardware replacement) for this even the CE is out of warranty as this is the result of following their instructions to do a software upgrade (which is confusing) to ACNS 5.

Anyone else have the same problem?

twaite · ‎10-26-2004

Well I have talked to two different techs who came to the conclusion that these need to be RMA'd. Athough you are right, they should have a solution in place, and they probaly do. Problem is the solution is just not for us end-users, meaning we are screwed with defective junk, or paying Cisco a large sum to repair poorly designed devices.

gid · ‎12-19-2004

FYI, what you are seeing is the flash image on a CE that has ECDN 2.x/3.x or ICDN software on it, not ACNS. The image you're seeing boot is supposed to be used only in the internal manufacturing procedure for those software loads and was never intended to be seen by end users. Most CE's -- ones running ACNS -- always boot from flash. This CE is supposed to have booted from disk, but for some reason it didn't, and it ends up booting the remnant mfg flash code. The reason this CE didn't boot from disk is unclear. Maybe the disk is bad or missing. Maybe the BIOS CMOS configuration area got corrupted and the BIOS reset to defaults.

Working around this situation -- basically getting the CE into a state where you can install ACNS -- would be fairly tricky. It's possible to load ACNS from a normal box running ECDN, but that requires that the CE boot normally from disk, which this CE is not doing.

beyondautomation · ‎12-19-2004

Gid,

Then please post a solution. No matter how tricky it is, we would like to try. There are many people outside stuck with unuseable CE's like this. Cisco, please provide a solution.

gid · ‎12-19-2004

I don't want to give the wrong impression -- recovery is far from guaranteed in this situation. It's an unfortunate characteristic of the old software load on this CE. At minimum the disk drive needs to be operational and needs to still contain the ECDN software.

Anyway, here is one thing to try: watch the console boot messages like you saw. Fairly early on when it says "Cisco CE Booting from Flash", there should be a delay of about 5 seconds. After this message is displayed, press H (I believe it will not be echoed back to you). This forces the BIOS to boot from disk. As above, if there is no usable SW on disk, then you are basically out of luck, unfortunately.

If you do get the CE to boot from disk, then re-try the ECDN to ACNS upgrade. (You'll probably have to go to ACNS 4.x first, then to ACNS 5.x, but if you get as far as having ACNS 4.x installed, then relatively speaking, you're in a much better position than you are now.) Watch the output on console carefully. I believe the ECDN upgrade image needs to do an intermediate reboot before it writes the ACNS image to flash. This means you'll probably have to do the press-H trick a second time. Anyway, what you really want to have happen is for ECDN to write an ACNS flash image -- if you can get ECDN to do that for you, then you'll be OK. But it's a big if, depending as I said on whether the disk-based software is still functional enough to boot.