Regulatory work?

Question

Documentation

Good afternoon!
Help to schedule maintenance work on switching equipment (switches, Cisco routers, optical media converters, DSLAM), servers, air conditioners. From the set of documents (instructions) for the equipment there was nothing that would describe the order, the frequency of the work performed (daily, monthly, annual maintenance).
I would like to find official documents (GOSTs and similar documents) that could be referred to when drawing up your plan of work carried out on the equipment.
ps For Soviet products (devices, communication equipment, etc.), complete with equipment, there was always a passport for the device, which indicated the characteristics, allowable values ​​during control measurements, the frequency and composition of preventive maintenance. I would like to find something like this for imported equipment, but it doesn’t work.

Answer 1

Well, it's like intelligent high-availability equipment, it is _usually_ supposed to work for years without maintenance.
Cisco.com has datasheets for their hardware. But it's a bit different there. In the case of servers, from the point of view of iron, their periodic shutdown for maintenance is also not provided.
What is the task? Usually, work is planned as needed (for example, a critical bug got out). Sometimes there are periodic update policies to fix security bugs, but this is also rare in the case of network hardware. Yes, and equipment with redundant soups can usually be updated without interrupting communication.

Answer 2

"on switching equipment ..." - they usually change after a failure or when upgrading the network, whichever comes first.
"Servers" - they themselves report, but if it died, then unfortunately without warning. Modernization, backups and redundancy ...
"air conditioners" - an inspection at least annually, and preferably twice a year. It is better to consult with a representative of the manufacturer’s company, but as a rule, it is better to clean outdoor units in early summer, after “everything and everything has faded”

Answer 3

If the manufacturer has not provided for the order of maintenance in the documentation, no one prevents the administrator from compiling it on his own, focusing on his own conditions and looking back at the main business of the company (so as not to interfere with the regulations). In general terms, the regulation is an external inspection, checking the rotation of all fans, if any, in a particular equipment, but monitoring does not track this, and external and internal cleaning of dust where it can get. See the place for the frequency and methods of conducting; it is advisable to turn off the equipment for cleaning from dust (if it is not just wiping the case, on which there are no ventilation holes). About kondei have already been written; servers can, for example, periodically test memory using memtest86 (although not a fact that is needed).

Answer 4

Thank you all, I myself adhere to the principle: if it works - do not touch it. Fans always need preventive maintenance ... this is as they become dirty.

Answer 5

Switches and routers are easiest to update during minimum load, and even better, update the hardware firmware, copy the configuration from the updated hardware, switch links, and send the current one to the firmware update and reserve, for example, if not a stack switch.
Servers are exactly the same - backup, update, restart services. if a cluster, then users should not notice anything at all. The updated server is put into maintenance mode and does not service connections.
if virtualka is even easier.