SCVJeff wrote:
You are talking about mechanical devices, none of which are internal to the router, so who cares?
No, really, power cycles do shorten lifespan. I've dealt with lots of equipment which runs constant on for years (like PBXs). The biggest fear with an extended power outage (they usually have battery and generator backup, but that only goes so far), or major upgrade which requires removing power is whether they'll come back up, and how many pieces which had been running fine for years won't power up anew.
It's really not debatable, power cycles produce more stress (thermal cycles, etc.) than no power cycles. They will reduce the lifespan. It's just a matter of whether the lifespan is shortened to less than the useful life of the product, many of which get replaced before they die on their own.
SCVJeff wrote:
I say again, IP controlled strips are common in the IT world to reset "junk" switch equipment from cheap Chinese junk companies like HP, Cisco, Dell, etc. maybe we should have them remove the word "enterprise class" from their description?
I've worked for years in data center environments - it's much more common to use PDUs to power cycle equipment because someone fat-fingered a configuration change than because the device just went dead on its own. That can happen, too, but it's a pretty rare occasion, certainly not daily or the device wouldn't be on the network. All non-trivial software has bugs. PDUs are also used to monitor power, and at times to shut off non-critical devices, such as during a power failure to conserve UPS capacity.
Having said that, I use a PDU to power cycle stuff, but only on rare occasion, not daily. Cable interface on the router doesn't come back up? The server will recognize the Internet is down, and power cycle it. Server goes down? I can console into the router and power cycle the server. Normally both stay running for long periods, it's mostly during poor power events (multiple power up/down after the UPS has run out, for example) that something does go wrong.
