3.4 Temperature
Temperature is often quoted as the most important envi-
ronmental factor affecting disk drive reliability. Previous
studies have indicated that temperature deltas as low as
15C can nearly double disk drive failure rates [4]. Here
we take temperature readings from the SMART records
every few minutes during the entire 9-month window
of observation and try to understand the correlation be-
tween temperature levels and failure rates.
We have aggregated temperature readings in several
different ways, including averages, maxima, fraction of
time spent above a given temperature value, number of
times a temperature threshold is crossed, and last tem-
perature before failure. Here we report data on averages
and note that other aggregation forms have shown sim-
ilar trends and and therefore suggest the same conclu-
sions.
We first look at the correlation between average tem-
perature during the observation period and failure. Fig-
ure 4 shows the distribution of drives with average tem-
perature in increments of one degree and the correspond-
ing annualized failure rates. The figure shows that fail-
ures do not increase when the average temperature in-
creases. In fact, there is a clear trend showing that lower
temperatures are associated with higher failure rates.
Only at very high temperatures is there a slight reversal
of this trend.
Figure 5 looks at the average temperatures for differ-
ent age groups. The distributions are in sync with Figure
4 showing a mostly flat failure rate at mid-range temper-
atures and a modest increase at the low end of the tem-
perature distribution. What stands out are the 3 and 4-
year old drives, where the trend for higher failures with
higher temperature is much more constant and also more
pronounced.
Overall our experiments can confirm previously re-
ported temperature effects only for the high end of our
temperature range and especially for older drives. In the
lower and middle temperature ranges, higher tempera-
tures are not associated with higher failure rates. This is
a fairly surprising result, which could indicate that data-
center or server designers have more freedom than pre-
viously thought when setting operating temperatures for
equipment that contains disk drives. We can conclude
that at moderate temperature ranges it is likely that there
are other effects which affect failure rates much more
strongly than temperatures do.

5 Conclusions
In this study we report on the failure characteristics of
consumer-grade disk drives. To our knowledge, the
study is unprecedented in that it uses a much larger
population size than has been previously reported and
presents a comprehensive analysis of the correlation be-
tween failures and several parameters that are believed to
affect disk lifetime. Such analysis is made possible by
a new highly parallel health data collection and analysis
infrastructure, and by the sheer size of our computing
deployment.
One of our key findings has been the lack of a con-
sistent pattern of higher failure rates for higher temper-
ature drives or for those drives at higher utilization lev-
els. Such correlations have been repeatedly highlighted
by previous studies, but we are unable to confirm them
by observing our population. Although our data do not
allow us to conclude that there is no such correlation,
it provides strong evidence to suggest that other effects
may be more prominent in affecting disk drive reliabil-
ity in the context of a professionally managed data center
deployment.
Our results confirm the findings of previous smaller
population studies that suggest that some of the SMART
parameters are well-correlated with higher failure prob-
abilities. We find, for example, that after their first scan
error, drives are 39 times more likely to fail within 60
days than drives with no such errors. First errors in re-
allocations, offline reallocations, and probational counts
are also strongly correlated to higher failure probabil-
ities. Despite those strong correlations, we find that
failure prediction models based on SMART parameters
alone are likely to be severely limited in their prediction
accuracy, given that a large fraction of our failed drives
have shown no SMART error signals whatsoever. This
result suggests that SMART models are more useful in
predicting trends for large aggregate populations than for
individual components. It also suggests that powerful
predictive models need to make use of signals beyond
those provided by SMART.

Полагаю неблагодарным интерпретировать каждую загогулину, всё ж безответственно привлеку внимание к тому, что AFR (годовой процент <былинного)))> отказа) на рис.4 для 50°C и 31°C в общем-то совпадает (!).