I have a M4100-50G-POE+ which has failed after a firmware update, the power led is blinking orange. Most manuals says it has failed POST, one of them says the power supply has failed - is it possible to temporarily power the unit (for testing only, no poe) through the RPS port if this is the case?
connecting to the CLI with a null modem cable I see the following before the unit re-initialises:
Initializing Arena. Initializing PCI. [normal] PCI[0]: Initializing host PCI[1]: Initializing host PCI[1] bus 0 slot 0/1: (24) not enough PCI mem space (134217728 requested) PCI[1] bus 0 slot 0/0: (24) not enough PCI mem space (134217728 requested) Initializing Devices.
Loader:stk Filesys:raw Dev:flash2.stk1 File: Options:(null) Loading: Could not load flash2.stk1:: Error Loader:stk Filesys:raw Dev:flash2.stk2 File: Options:(null) Loading: Could not load flash2.stk2:: Error **Exception 32: EPC=87E8A6D8, Cause=0000001C (BusErrD ) RA=87E8A6D4, VAddr=B54F3AC0
Initializing Arena. Initializing PCI. [normal] PCI[0]: Initializing host PCI[1]: Initializing host PCI[1] bus 0 slot 0/1: (24) not enough PCI mem space (134217728 requested) PCI[1] bus 0 slot 0/0: (24) not enough PCI mem space (134217728 requested) Initializing Devices.
Can anyone tell me if this is unit is recoverable or should I start the warranty process? Is it possible to reload the correct firmware? (or is this a hardware failure?)
Thank you
Wayne
Model: M4100-50G-PoE+ (GSM7248Pv1h1)|46‐port GE + 4 GE Combo L2 Managed PoE+ Switch
As I checked the initialization process of the switch that it can't load the flash2.stk1. This is a bad firmware upgrade result that it cannot load the image from the flash. May I know what firmware version are you trying to upload? Please make sure that your are uploading the correct firmware for your M4100 device. You may still recover the switch by reflashing the switch via CLI using XModem.
Thank you for your reply and the link to the recovery procedure.
Looking at my CLI output it appears that the switch has no access to its flash memory (so I never get to "Starting program at 0xXXXXXXXX" and cannot stop the boot sequence and enter the boot menu in the link provided).
It all seems to go astray at the line starting "Loading:"
The example in your link continues with "STK image is 12324100 bytes, CRC 0xFC98 3605504 bytes"
The switch I am working on returns " Could not load flash2.stk1:: Error"
Prior to that there are also "not enough PCI mem space" errors (lines starting PCI)
However I can get to the CFE> prompt by pressing Ctrl-C, is there another recovery procedure I can use?
Thank you for your reply, as I said above I cannot get into the CLI menu (by pressing and holding q), please see Appendix 1 below for the console output.
Thanks also for letting me know the link supplied was only for your routers. Could you please let me know if there is a recovery procedure for your switches? (when the Netgear CLI is inaccessible) If not, I will return the switch under warranty - it is only 1 month old and its POE did not function correctly. Updating the firmware, as recommended by Netgear support left the switch in this un-usable state.
As I also pointed out above, I can get into the CFE console by pressing Control-C. I am guessing you are using Broadcom hardware in your switches and this is their firmware interface/bootloader. Please let me know if I can use this to reload the netgear firmware. If not, I will return the swiitch under warranty.
Thank you
Wayne
Appendix 1 (console output of not being able to access the CLI menu by pressing and holding the q key):
qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqq
Initializing Arena. Initializing PCI. [normal] qPCI[0]: Initializing host qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqPCI[1]: Initializing host qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqPCI[1] bus 0 slot 0/1: (24) not eno ugh PCIq mem space (134217728 requested) PCI[1] bus 0 slot 0/0: (24) not enough PCI mem space (134217728 requested) Initializing Devices.
It seems that you already tried logging in to the boot menu and still doesn't work. Kindly open a chat / case online on NETGEAR support and please let them know the status of the switch. Also be ready with a soft copy of the Proof of Purchase or Invoice of the device for warranty verification.
I had a firmware updates fail when upgrading several units when going from 10.0.1.16 to 10.0.2.26. Exact boot console conditions as previously stated. Switch was effectively dead as it rebooted itself. I made the mistake of trying image1 and image2... that's when I lost functionality of the switch completely. I noted on other units where the firmware upgrade failed, that image1 I regularly tried to upgrade was blank or gone when the switch rebooted on it's own after an attempt. (checked this in Maintenance, File Management...Dual Image Configuration) The image2 was the only one allowing the switch to boot up... until I wiped it out too during an attempt to try flashing image2. Looked all over for a recovery process and was able to get to a CFE prompt with Ctrl-C and not much else. Not able to access the typical recovery options. I submitted an RMA for a replacement, that won't get here till Monday(being Friday). Since I had the unit yet, I decided to see what I could do in the CFE arena. Did find a number of sample interactions in the CFE CLI, none for the netgear switch itself, but did get enough incite to remedy the situation. Supprised Netgear doesn't post the recovery procedure for a semi-bricked unit that is in this state. It wasn't too difficult. Now I don't need the RMA'd replacement unit anymore.
Basic steps were to get into CFE... easy enough... reactivate eth0 with the -auto switch manually so the switch could see my network for tftp... used the flash command to overwrite the flash2.stk1 image via tftp. It rebooted and all was good again.
