Juniper M5/M10 Midplane problems ??
John L Lee
johnllee at mindspring.com
Mon Jan 6 16:12:11 UTC 2003
john,
While I/we have tested the M5/M10 series from Juniper we did not role
them out in the field. The Mid plane design
is / was a standard design to seperate packet, cell, or tdm processing
from the electrical or optical interfaces. You would have to
replace the interface card and cables more often than the processing
modules and you can have the same processing module
drive one or more interface cards.
Usually the mid plane is passive or only has passive (resistor,
capacitor, etc..) components on it for signal enhancement. If the board
insertion and extraction hardware works than bent pins are supposed to
be eliminated or at least minimized. Some designs allow you to insert
the board and then when the pins are lined up apply pressure to
positively hold them together. I am not as familiar with the "new" high
denisty board interconnect hardware but since tolorances are tighter
they have to address alighment, insertion and extraction issues to a
greater degree.
In the lab enviroment we swapped cards many more times than you would in
an operational system and did not have a
problem. I presume that the cards are hot swappable ( we tested alot of
hardware) and if that is the case then there needs to be hardware to
ground before circuit contact to not allow floating voltages to destroy
the cards cmos and other static sensitive electronics.
From a Service perspective I would reccomend one or more spare chassis
and or fully loaded and configured systems for maintenance. The
diagnostics usually allow you to trouble shoot to the board level or
power supply level and replace the FRU (field replaceable unit), but to
improve mean time to repair I would swap out and I would not do a major
repair like a midplane replacement in a POP without space and proper
test equipment.
john (I Still Don't kNow) lee
john at chagresventures.com wrote:
>Been looking at a M5/M10 box and have noticed its
>not real easy (unless I'm blind) to replace the
>mid-plane.
>
>Thus I'm wondering what people in the field have
>experienced with respect to mid-plane failures.
>
>1. Bent or broken pins on the PIC side
>
>2. Circuit failures
>
>
>My eval of this product is for a client and their
>infrastructure needs.
>
>
>Other bits of learned wisdom also excepted and
>encouraged :)
>
>Thanks
>
>john brown
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.nanog.org/pipermail/nanog/attachments/20030106/fa791912/attachment.html>
More information about the NANOG
mailing list