<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
(	O
SRE	O
)	O
is	O
a	O
set	O
of	O
principles	O
and	O
practices	O
that	O
incorporates	O
aspects	O
of	O
software	B-General_Concept
engineering	I-General_Concept
and	O
applies	O
them	O
to	O
IT	B-Application
infrastructure	I-Application
and	O
operations	O
.	O
</s>
<s>
The	O
main	O
objectives	O
are	O
to	O
create	O
highly	O
reliable	O
and	O
scalable	B-Architecture
software	O
systems	O
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
has	O
been	O
described	O
as	O
a	O
specific	O
implementation	O
of	O
DevOps	O
.	O
</s>
<s>
The	O
field	O
of	O
site	B-Application
reliability	I-Application
engineering	I-Application
originated	O
at	O
Google	B-Application
with	O
Ben	O
Treynor	O
Sloss	O
,	O
who	O
founded	O
a	O
site	O
reliability	O
team	O
after	O
joining	O
the	O
company	O
in	O
2003	O
.	O
</s>
<s>
In	O
2016	O
,	O
Google	B-Application
employed	O
more	O
than	O
1,000	O
site	B-Application
reliability	I-Application
engineers	I-Application
.	O
</s>
<s>
After	O
originating	O
at	O
Google	B-Application
in	O
2003	O
,	O
the	O
concept	O
spread	O
into	O
the	O
broader	O
software	O
development	O
industry	O
,	O
and	O
other	O
companies	O
subsequently	O
began	O
to	O
employ	O
site	B-Application
reliability	I-Application
engineers	I-Application
.	O
</s>
<s>
Organizations	O
who	O
have	O
adopted	O
the	O
concept	O
include	O
Airbnb	O
,	O
Dropbox	B-Application
,	O
IBM	O
,	O
LinkedIn	B-Application
,	O
Netflix	O
,	O
and	O
Wikimedia	O
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
,	O
as	O
a	O
job	O
role	O
,	O
may	O
be	O
performed	O
by	O
solo	O
practitioners	O
or	O
organized	O
in	O
teams	O
usually	O
being	O
responsible	O
for	O
a	O
combination	O
of	O
the	O
following	O
within	O
a	O
broader	O
engineering	O
organization	O
:	O
System	O
availability	O
,	O
latency	O
,	O
performance	O
,	O
efficiency	O
,	O
change	O
management	O
,	O
monitoring	B-Protocol
,	O
emergency	O
response	O
,	O
and	O
capacity	O
planning	O
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineers	I-Application
often	O
have	O
backgrounds	O
in	O
software	B-General_Concept
engineering	I-General_Concept
,	O
system	O
engineering	O
,	O
or	O
system	B-Application
administration	I-Application
.	O
</s>
<s>
Focuses	O
of	O
site	B-Application
reliability	I-Application
engineering	I-Application
include	O
automation	O
,	O
system	O
design	O
,	O
and	O
improvements	O
to	O
system	B-Application
resilience	I-Application
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
,	O
as	O
a	O
set	O
of	O
principles	O
and	O
practices	O
,	O
can	O
be	O
performed	O
by	O
anyone	O
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
has	O
also	O
been	O
described	O
as	O
a	O
specific	O
implementation	O
of	O
DevOps	O
,	O
but	O
it	O
focuses	O
specifically	O
on	O
building	O
reliable	O
systems	O
,	O
whereas	O
DevOps	O
is	O
more	O
broadly	O
focused	O
.	O
</s>
<s>
There	O
have	O
been	O
multiple	O
attempts	O
to	O
define	O
a	O
canonical	O
list	O
of	O
site	B-Application
reliability	I-Application
engineering	I-Application
principles	O
,	O
but	O
while	O
consensus	O
is	O
lacking	O
,	O
the	O
following	O
characteristics	O
are	O
usually	O
included	O
in	O
most	O
definitions	O
:	O
</s>
<s>
Observability	B-Operating_System
,	O
as	O
in	O
,	O
the	O
ability	O
to	O
be	O
able	O
to	O
ask	O
arbitrary	O
questions	O
about	O
your	O
system	O
without	O
having	O
to	O
know	O
ahead	O
of	O
time	O
what	O
you	O
wanted	O
to	O
ask	O
.	O
</s>
<s>
The	O
site	B-Application
reliability	I-Application
engineering	I-Application
practices	O
also	O
vary	O
widely	O
,	O
but	O
the	O
list	O
below	O
is	O
relatively	O
commonly	O
seen	O
being	O
at	O
least	O
partially	O
implemented	O
:	O
</s>
<s>
Designing	O
for	O
and	O
implementing	O
observability	B-Operating_System
.	O
</s>
<s>
Chaos	B-Application
engineering	I-Application
.	O
</s>
<s>
Site	B-Application
reliability	I-Application
engineering	I-Application
teams	O
engage	O
with	O
the	O
other	O
teams	O
within	O
their	O
companies	O
and	O
the	O
SRE	O
principles	O
and	O
practices	O
in	O
various	O
forms	O
.	O
</s>
<s>
Infrastructure	B-Application
SRE	O
teams	O
may	O
pair	O
up	O
with	O
one	O
or	O
more	O
platform	O
engineering	O
team(s )	O
,	O
but	O
they	O
differ	O
in	O
that	O
Infrastructure	B-Application
SRE	O
teams	O
focuses	O
on	O
performing	O
most	O
,	O
if	O
not	O
all	O
,	O
of	O
the	O
work	O
described	O
in	O
the	O
principles	O
and	O
practices	O
list	O
above	O
.	O
</s>
<s>
Usually	O
SRE	O
solo	O
practitioners	O
or	O
pairs	O
staffed	O
within	O
a	O
software	B-General_Concept
engineering	I-General_Concept
team	O
to	O
apply	O
most	O
of	O
the	O
principles	O
and	O
practices	O
described	O
above	O
.	O
</s>
<s>
multiple	O
Product/application	O
SRE	O
teams	O
to	O
meet	O
specific	O
demands	O
of	O
several	O
products	O
and	O
an	O
Infrastructure	B-Application
SRE	O
team	O
to	O
pair	O
up	O
with	O
a	O
Platform	O
engineering	O
group	O
to	O
meet	O
reliability	O
goals	O
of	O
a	O
common	O
platform	O
for	O
both	O
products/applications	O
.	O
</s>
<s>
The	O
USENIX	B-Operating_System
organization	O
has	O
held	O
an	O
annual	O
SREcon	O
conference	O
since	O
2014	O
for	O
site	B-Application
reliability	I-Application
engineers	I-Application
in	O
the	O
industry	O
,	O
and	O
also	O
holds	O
regional	O
conferences	O
with	O
similar	O
themes	O
.	O
</s>
