Dear customers!

Please tell me about MTBF (mean time between failures, a measure of the reliability of a device or system) for RouterBOARD hardware. Where can I get this information?

Great Thnx!

P.S. It would be great if you could send some sort of a document confirming this...

From my knowledge MikroTik doesn't have such data (because no official tests are preformed to determine MTBF).
然而从我观察你可以预期around 4-5 years lifetime (for 24x7 scenario). In my opinion 3 years is a sensible rotation period due to technology evolution.

From my knowledge MikroTik doesn't have such data (because no official tests are preformed to determine MTBF).
然而从我观察你可以预期around 4-5 years lifetime (for 24x7 scenario). In my opinion 3 years is a sensible rotation period due to technology evolution.

This heavily depends. Some boards never die some very early. 1xx/5xx was undestroyable. 600 had weak ports. SXT tend to die earlier.
3 Years is to short. E.g. we have RB1000 we dont want to replace.

Imho since in RB hardware there are no moving parts I will expect higher mtbf times than a mechanical HD drive.

"Hard drive manufacturers quote average hard drive life as Mean Time Between Failures (MTBF). Ordinary consumer drives are in the 500,000 hour MTBF range, while enterprise (e.g., more expensive) hard drives can range up to 1.5 million hour MTBF. Since there are 8,760 hours in a 24x7 year, 500,000 hour average MTBF is a lot of years. Right? Yeah, 57 years is the answer"

Thnx!
I really hope to see the official comments of developers.

We have not made any MTBF calculations for any of our devices. These calculations are purely theoretical, they do not reflect real life, so we don't provide them for our products.

We replace them 3-4 years tops.

After that you deal with capacitor , power supply and flash issues down the line.

We have not made any MTBF calculations for any of our devices. These calculations are purely theoretical, they do not reflect real life, so we don't provide them for our products.

No. If you track the failure rate/returns of your products you will see very different numbers between your products. This is no theoretical value.

The selection of components heavily influence the live time. So you might categorize your products in "normal user" / "industrial" / "rugged/extended life time". Esp. the capacitors in outdoor devices or the selection of UV resistent plastics might be a difference.
This would give us the chance to select the right product for the use case. E.g. we give "normal" or "business grade" CPEs to our users. "Normal" is a SXT Lite. "business grade" is a Mars Antenna with metal enclosure and .ac board inside.

We see SXTs coming back with popped up Ethernetport or some dont boot anymore. Some drop the wireless connection from time to time. The plastic moves to yellow after a short time in the sun. So this is the cheap line. This is normal. We buy a cheap CPE so we get cheaper components. But when a user/we want to spend more money to get a more rugged device we need an indicator what board/device to buy. At the moment we talk to the distributor to get an idea ...

We have not made any MTBF calculations for any of our devices. These calculations are purely theoretical, they do not reflect real life, so we don't provide them for our products.

Strongly disagree! This is the most important characteristic! This allows you to determine the degree of reliability and applicability to large and serious projects.

I remember RB100 and 500 series. Some of them still in use after 5 years of time.
Newer products tend to fail earlier than that.
I wish we could find the same stability with the new products compared to the old series.

We always use meanwell power supply's with batt pack

in 10 jears time on a network with 5000 wireless devies and 800 mikrotiks and we had like 12 mikrotiks broken for no reason. , at the moment we use mikrotik for routing, 87% of the wireless hardware is from a other brand.

and +- 40 mikrotiks did go broken beqause of water, lightning, .... we replace never old hardware , only if its realy old and if the site is growing.
if we replace a backbone router , then it will be re-used on a end point.

We use RB493, omnitik, RB2011, Groove, CCR, CRS, RB433, RB951, ..... the omnitik does broke more often than the 493 with lightning, in this case 1 or 2 ethernet interfaces stop working.

So if the powersupply is good , you will never have problems with mikrotik.
from our 400 meanwell power supply's there untill now not one broken , even after a lightning storm.
fake meanwell power supply's (other brand, same shape) do go broken in 2-3 years

In my experience I never had troubles with high end routers (I have still the first RB1100 model in production). I have also lots of RB750 (around 500) installed since ~ 2009, and only a few dozens have been replaced so far (often because of external events, like surges on the ethernet interfaces).

That said, if history repeats itself, you can expect those to live at least 5 years.

MTBF is a syntetic calculation and is not comparable to real life observation. Of course in real life "some boards" die sooner as you wrote above. MTBF calculation does not take this into account. It simply sums the expected lifetimes of the parts used, adds some statistical variables and calculates an unrealistic number that is not helpful for real life. This is why we don't spend a lot of time calculating all the values.

I emailed MikroTik for the MTBF of the RB450G. Their response was 120,000 hours, using this Mil-spec handbook - "Reliability Prediction of Electronic Equipment MIL-HDBK-217F"

i still have a rb450G in use for years 24/7. The main problems i have with routerboards are simply the PSU, they break/fail too easily other than that the routerboard themselves are good. Mikrotik says they have another company that makes their PSUs, i hope they arent using them anymore as i had to replace my CCR's PSU twice, the first replacement i bought was one usually used for network related installs like cameras and other equipment usually tucked away as part of the structure, the 2nd replacement is a laptop style PSU which is working really well but i had to get a DC power jack and solder the wires onto them and glue it to the insulated part of the case.

UP!
Mikrotik APs compliant with the wifi4eu minimum specs?

As request fromWiFi4EU9.2.1 What are the technical requirements for the WiFi4EU Access Points?

The technical specifications of the equipment are detailed in section 6.2.2 of the call text, as well as Article I.2 of Annex I of the Grant Agreement signed between the municipalities and the Commission.

The municipality shall ensure that each Access Point:

Supports concurrent dual-band (2,4Ghz – 5Ghz) use
Has a support cycle superior to 5 years
Has a mean time between failure (MTBF) of at least 5 years
Has a dedicated and centralised single point of management for all APs of each WiFi4EU network
Supports IEEE 802.1x
Complies with IEEE 802.11ac Wave I
Supports IEEE 802.11r
Supports IEEE 802.11k
Supports IEEE 802.11v
Is able to handle at least 50 concurrent users without performance degradation
Has at least 2x2 multiple-input-multiple-output (MIMO)
Complies with Hotspot 2.0 (Passpoint Wi-Fi Alliance certification program).


There is avaiable information for MTBF?

UP!
Mikrotik APs compliant with the wifi4eu minimum specs?

As request fromWiFi4EU9.2.1 What are the technical requirements for the WiFi4EU Access Points?
(...)
Supports IEEE 802.11r
Supports IEEE 802.11k
Supports IEEE 802.11v
(...)

These protocols are missing in Mikrotik products, so they are not compliant anyway.