BFT failure or HDD failure?

[kpzan]

Hello gentlemen, hope everyone is doing well.

Got a question about a BCM450 6.0 system -

When a BFT fails, how do the symptoms differ from when an HDD fails?

How does a technician know it's the BFT and not the HDD? Or vice versa?

Got a customer with a BCM450 6.0 that's been acting erratic and just trying to prepare myself in case the system crashes.

Thanks in advance.

 

[exsmogger]

It's likely the hard drive. I've not seen a bad BFT on a BCM450 in all my years. The only way to know would be to replace the hard drive and see if it does the same things.

I have seen flaky power connections behind the MBM slots, and bad power supplies, so those are other things to check.

Give the alarm log a look when you first get there. It might lead you toward the problem. Good luck!

Brian Cox
Georgia Telephone

http://Georgia-Telephone.com

http://www.linkedin.com/in/briancox1952

https://brian-cox.square.site/

 

[ucxguy]

I'd also say that the most likely component to fail is either the hard drive or the power supply. Could you expand on the "acting erratic" a bit and describe the symptoms?

 

[kpzan]

Thanks exsmooger. I'll check the alarm log as you advise.

ucxguy - the customer tells me that once every couple of days all the phones will "freeze up" where none of the buttons work on any of the phones. Any call in progress is lost. Indicator lights and the time and date are also "frozen". The customer will power cycle the BCM and it will come back on as normal.

The system has around 60 T7316Es and a PRI.

 

[exsmogger]

Another thought is that system resources might be getting low. A CEC card is needed to add more resources when going over 48 trunks or 100 stations of any type. I don't know if the alarm log will reflect this problem, but that's where I would start.

Brian Cox
Georgia Telephone

http://Georgia-Telephone.com

http://www.linkedin.com/in/briancox1952

https://brian-cox.square.site/

 

[ucxguy]

I don't think the DSP resources are the problem here - if the system is out of DSP resources, additional calls would not be established, but everything else would work normally. So, the CEC won't solve it. You could run BCM Monitor to see the DSP usage.

If everything suddenly "freezes", the reason usually is that the Linux OS stopped working (due to "kernel panic", lockup of an I/O bus or something similar) and with that all applications that were running. It is not simple to determine where the actual problem is. It could be power related (so you could try another power supply), it could be a bad memory stick or it could be some corrupt executable being read from the hard drive.

 

[kpzan]

Thank you both for your technical expertise and kind advice.

If the system continues to do this, I'll replace the power supply. If replacing the power supply doesn't clear the trouble, I'll try replacing the hard drive.

 

[LXLS]

While I agree that it's most likely to be the HDD, if you want to do some other testing, here's what I'd do before ruling it as a hardware problem:

Boot the machine to a Memtest86 diskette/USB stick/CD/whatever and let it do a few passes of the memory that's in the box. Have you considered converting this machine over to SSD as well? What kind of environment does the BCM live in? It is also likely time for a CPU repaste, especially if it lives in a dirty/hot environment.