MSA1000 San Switch

[Ernest R. Reed, Jr.]

While logged into the San Switch it continually interrupts with "Committing configuration...done." followed by "ë?¡"the the next line "¹?¡", this line continues until a 1 appears and switch is no longer accessible. I close the connection wait a couple of seconds then can access it again until this process repeats itself.
On occasion I have gotten "After Firmware Download it is highly recommended t" I am assuming the rest is "to reboot the switch".
Figuring there was an issue when the switch firmware was updated I attempted to redo it. But unlike other switches I've done it returns a firmware not found. Here is my process, which I just did on another switch and verified it works:
MSA1000-switch1:admin> firmwaredownload
Server Name or IP Address [host]: 10.11.xxx.xx
User Name [user]: xxx
File Name [/usr/switch/firmware]: v3.2.1c
Protocol (RSHD or FTP) [rshd]: FTP
Password:
File not found.
MSA1000-switch1:admin>
(I replaced the IP with x's for security and x'd out the user name but they are correct on the real attempt)

Switch Bad???

[Pieter 't Hart]

>>> While logged into the San Switch <<<
is this on the console port of the switch, or through the network?

If console port, check network connectivity between the switch and 10.11.xxx.xx
eg from 10.11.xxx.xx try to telnet/ssh/http to the switch.

From another host/PC try "ftp 10.11.xxx.xx " using the same account
in this ftp session :
- do a "ls" for the default directory
- and "cd /usr/switch/firmware" + "ls"
to check if "v3.2.1c" exists there and is accessible.

[Ernest R. Reed, Jr.]

I am logging in via a network connection, and the Switch is at v3.2.1c. I have accessed it through a number of different methods, including direct connection from a laptop serial to the switch serial and always get the same "ë?¡"the the next line "¹?¡" and then accessibility freezes. I am currently working it stone age style. I have a laptop configured as an FTP Server, it already has the current v3.2.1c firmware on it. I access the switch via telnet using reflections to connect to a DECServer and accessing the switch by connecting to the port it is assigned. I can log into the switch and I can give it commands which it responds to, somewhere in this process it starts sending back "Committing configuration...done" but nothing was done and no changes were made. I got this right after logging onto the switch today. Did a switchshow and the switch returned the following:

MSA1000-switch1:admin> switchShow
switchName: MSA1000-switch1
switchType: 18.2
switchState: Online
switchMode: Native
switchRole: Principal
switchDomain: 1
switchId: fffc01
switchWwn: 10:00:00:60:69:ca:a8:d6
switchBeacon: OFF
Zoning: OFF
port 0: cu 2G Online F-Port 50:08:05:f3:00:0f:b9:31
port 1: id 2G Online F-Port 10:00:00:00:c9:2a:79:bf
port 2: id 2G No_Light
port 3: id N2 No_Light
port 4: id N2 No_Light
port 5: -- N2 No_Module
port 6: -- N2 No_Module
port 7: -- N2 No_Module

It is obvious the switch is working but how well and why it is constantly committing configuration or why it is throwing the garbage and freezing up is what I'm attempting to determine.

[Pieter 't Hart]

>>> I access the switch via telnet using reflections to connect to a DECServer and accessing the switch by connecting to the port it is assigned. <<<

this way you do NOT access the switch over the network!
You access the DECserver over the network and access the switch over it's serial console.

=> check again if you have network connectivity (directly) between laptop and switch.
what ip-address does the switch has?
can you ping this address from your laptop.

[Pieter 't Hart]

use "ipAddrShow" to retreive the switch's ip-address

[Ernest R. Reed, Jr.]

I can connect to the switch using any and all options described. It does not matter how I access it each method ends with the same result. Within minutes of connecting I start receiving a "Committing configuration...done." and then get the garbage mentioned in earlier strings. I spoke with other Engineers and all mentioned that on occasion they too have seen this and while it appears the switch is functioning properly it is actually failing over to the other switch, it does not loose fiber connectivity so push come to shove it can continue to remain in the MSA1000 until a replacement San Switch is received. I did finally get to re-install the v3.2.1c firmware and it did not make a difference so I am resolved to the fact the switch is indeed defective. I was hoping this was not the case or there was something someone had done in the past that fixed this issue but it appears mine is the only case posted.
Thanks for your time, effort, and suggestions working this out!

[Pieter 't Hart]

Is it an option to
- remove all fiber cables
- reset the switch to it's factory default configuration
- hard reset the swith by pulling it from the chassis (and remounting)
- check if problem remains
= if no problems
- reload the switch configuration
- check if problem returns
= if problem returns, repeat reset, load config step by step untill offending config item is found.
= if stable
- reconnect the fibre cables.

[Ernest R. Reed, Jr.]

You must be reading my mind. I have a testing platform that I put the switch in for my troubleshooting, we are a test facility and have the resources to test and check parts on or off line the environment. I did the things, suggested, yesterday and still no change. In fact it took seven attempts to get the switch reset to default. Each time I attempted it it returned configuration change done but when I'd do a showSwitch I saw the same configuration I started with. The other thing I found out performing these processes is that when the switch freezes up it silently fails over to the other one, there is no indication when it does. I found out because I removed the second switch and it could not fail over so my MSA1000 told me it had a catastrophic hardware failure.
This morning I received another switch that is doing that exact same thing and I am getting the exact same results. I even put both switches together in the MSA to see what would happen, nothing happened MSA reported the same catastrophic hardware failure. So both switches are bad.