Monitorama 2015 Netflix Instance Analysis

Netflix
Instance Performance Analysis
Requirements
Brendan Gregg
Senior Performance Architect
Performance Engineering Team
bgregg@netflix.com @brendangregg
Jun
2015

Monitoring companies are selling
faster horses
I want to buy a car

Server/Instance Analysis Potential
In the last 10 years…
•  More Linux
•  More Linux metrics
•  Better visualizations
•  Containers
Conditions ripe for innovation: where is our Henry Ford?

This Talk
•  Instance analysis: system resources, kernel, processes
–  For customers: what you can ask for
–  For vendors: our desirables & requirements
–  What we are building (and open sourcing) at Netflix to
modernize instance performance analysis (Vector, …)

•  Over 60M subscribers
•  FreeBSD CDN for content delivery
•  Massive AWS EC2 Linux cloud
•  Many monitoring/analysis tools
•  Awesome place to work

Agenda
1.  Desirables
2.  Undesirables
3.  Requirements
4.  Methodologies
5.  Our Tools

Histograms
…
or
a
density
plot

Monitorama 2015 Netflix Instance Analysis

Other Desirables
•  Safe for production use
•  Easy to use: self service
•  [Near] Real Time
•  Ad hoc / custom instrumentation
•  Complete documentation
•  Graph labels and units
•  Open source
•  Community

Tachometers
…especially with arbitrary color highlighting

Pie Charts
…for real-time metrics
usr
sys
wait
idle

Doughnuts
usr
sys
wait
idle

…like pie charts but worse

Traffic Lights
…when used for subjective metrics
These can be used for objective metrics
For subjective metrics (eg, IOPS/latency) try weather icons instead
RED == BAD (usually)
GREEN == GOOD (hopefully)

Acceptable T&Cs
•  Probably acceptable:
•  Probably not acceptable:
•  Check with your legal team
By
submi9ng
any
Ideas,
Customer
and
Authorized
Users
agree

that:
...
(iii)
all
right,
Ftle
and
interest
in
and
to
the
Ideas,
including
all

associated
IP
Rights,
shall
be,
and
hereby
are,
assigned
to
[us]

XXX,
Inc.
shall
have
a
royalty-‐free,
worldwide,
transferable,
and

perpetual
license
to
use
or
incorporate
into
the
Service
any

suggesFons,
ideas,
enhancement
requests,
feedback,
or
other

informaFon
provided
by
you
or
any
Authorized
User
relaFng
to
the

Service.

Acceptable Technical Debt
•  It must be worth the …
•  Extra complexity when debugging
•  Time to explain to others
•  Production reliability risk
•  Security risk
•  There is no such thing as a free trial

Known Overhead
•  Overhead must be known to be managed
–  T&Cs should not prohibit its measurement or publication
•  Sources of overhead:
–  CPU cycles
–  File system I/O
–  Network I/O
–  Installed software size
•  We will measure it

Low Overhead
•  Overhead should also be the lowest possible
–  1% CPU overhead means 1% more instances, and $$$
•  Things we try to avoid
–  Tracing every function/method call
–  Needless kernel/user data transfers
–  strace (ptrace), tcpdump, libpcap, …
•  Event logging doesn't scale

Scalable
•  Can the product scale to (say) 100,000 instances?
–  Atlas, our cloud-wide analysis tool, can
–  We tend to kill other monitoring tools that attempt this
•  Real-time dashboards showing all instances:
–  How does that work? Can it scale to 1k? … 100k?
–  Adrian Cockcroft's spigo can simulate protocols at scale
•  High overhead might be worth it: on-demand only

Useful
An instance analysis solution must provide
actionable information
that helps us improve performance

Methodologies
Methodologies pose the questions
for metrics to answer
Good monitoring/analysis tools should support
performance analysis methodologies

Drunk Man Anti-Method
•  Tune things at random until the problem goes away

Workload Characterization
Study the workload applied:
1.  Who
2.  Why
3.  What
4.  How
Target
Workload

Workload Characterization
Eg, for CPUs:
1.  Who: which PIDs, programs, users
2.  Why: code paths, context
3.  What: CPU instructions, cycles
4.  How: changing over time
Target
Workload

CPUs
Who
How What
Why
top,
htop!
perf record -g!
ﬂame
graphs

monitoring
perf stat -a -d!

Most Monitoring Products Today
Who
How What
Why
top,
htop!
perf record -g!
ﬂame
Graphs

monitoring
perf stat -a -d!

The USE Method
•  For every resource, check:
1.  Utilization
2.  Saturation
3.  Errors
•  Saturation is queue length or queued time
•  Start by drawing a functional (block) diagram of your
system / software / environment
Resource

UFlizaFon

(%)
X

USE Method for Hardware
Include busses & interconnects!

hXp://www.brendangregg.com/USEmethod/use-‐linux.html

Most Monitoring Products Today
•  Showing what is and is not commonly measured
•  Score: 8 out of 33 (24%)
•  We can do better… U
S
E

U
S
E

U
S
E

U
S
E

U
S
E

U
S
E

U
S
E

U
S
E
U
S
E
U
S
E
U
S
E

Other Methodologies
•  There are many more:
–  Drill-Down Analysis Method
–  Time Division Method
–  Stack Profile Method
–  Off-CPU Analysis
–  …
–  I've covered these in previous talks & books

BaseAMI
•  Many sources for instance metrics & analysis
–  Atlas, Vector, sar, perf-tools (ftrace, perf_events), …
•  Currently not using 3rd party monitoring vendor tools
Linux
(usually
Ubuntu)

Java
(JDK
7
or
8)

Tomcat
GC
and

thread

dump

logging

hystrix,
metrics
(Servo),

health
check

OpFonal
Apache,

memcached,
Node.js,

…

Atlas,
S3
log
rotaFon,

sar,
erace,
perf,
stap,

perf-‐tools

Vector,
pcp

ApplicaFon
war
ﬁles,

plahorm,
base
servelet

Netflix Atlas
Select
Instance

Historical
Metrics

Select
Metrics

Netflix Vector
Near
real-‐7me,

per-‐second
metrics

Flame
Graphs

Select

Metrics

Select
Instance

Needs -XX:+PreserveFramePointer
and perf-map-agent
Java CPU Flame Graphs
Java
JVM

Kernel

sar
•  System Activity Reporter. Archive of metrics, eg:
•  Metrics are also in Atlas and Vector
•  Linux sar is well designed: units, groups
$ sar -n DEV!
Linux 3.13.0-49-generic (prod0141) !06/06/2015!_x86_64_ !(16 CPU)!
!
12:00:01 AM IFACE rxpck/s txpck/s rxkB/s txkB/s rxcmp/s txcmp/s rxmcst/s %ifutil!
12:05:01 AM eth0 4824.26 3941.37 919.57 15706.14 0.00 0.00 0.00 0.00!
12:05:01 AM lo 23913.29 23913.29 17677.23 17677.23 0.00 0.00 0.00 0.00!
12:15:01 AM eth0 4507.22 3749.46 909.03 12481.74 0.00 0.00 0.00 0.00!
12:15:01 AM lo 23456.94 23456.94 14424.28 14424.28 0.00 0.00 0.00 0.00!
12:25:01 AM eth0 10372.37 9990.59 1219.22 27788.19 0.00 0.00 0.00 0.00!
12:25:01 AM lo 25725.15 25725.15 29372.20 29372.20 0.00 0.00 0.00 0.00!
12:35:01 AM eth0 4729.53 3899.14 914.74 12773.97 0.00 0.00 0.00 0.00!
12:35:01 AM lo 23943.61 23943.61 14740.62 14740.62 0.00 0.00 0.00 0.00!
[…]!

perf-tools
•  Some front-ends to Linux ftrace & perf_events
–  Advanced, custom kernel observability when needed (rare)
–  https://github.com/brendangregg/perf-tools
–  Unsupported hacks: see WARNINGs
•  ftrace
–  First added to Linux 2.6.27
–  A collection of capabilities, used via /sys/kernel/debug/tracing/
•  perf_events
–  First added to Linux 2.6.31
–  Tracer/profiler multi-tool, used via "perf" command

perf-tools: funccount
•  Eg, count a kernel function call rate:
•  Other perf-tools can then instrument these in more detail
# ./funccount -i 1 'bio_*'!
Tracing "bio_*"... Ctrl-C to end.!
!
FUNC COUNT!
bio_attempt_back_merge 26!
bio_get_nr_vecs 361!
bio_alloc 536!
bio_alloc_bioset 536!
bio_endio 536!
bio_free 536!
bio_fs_destructor 536!
bio_init 536!
bio_integrity_enabled 536!
bio_put 729!
bio_add_page 1004!
!
[...]!
Counts
are
in-‐kernel,

for
low
overhead

eBPF
•  Currently being integrated. Efficient (JIT) in-kernel maps.
•  Measure latency, heat maps, …

eBPF
eBPF will make a profound difference to
monitoring on Linux systems
There will be an arms race to support it, post Linux 4.1+
If it's not on your roadmap, it should be

Requirements
•  Acceptable T&Cs
•  Acceptable technical debt
•  Known overhead
•  Low overhead
•  Scalable
•  Useful

Methodologies
Support for:
•  Workload Characterization
•  The USE Method
•  …
Not starting with metrics in search of uses

Instrument These
With full eBPF support
Linux has awesome instrumentation: use it!

Links & References
•  Netflix Vector
–  https://github.com/netflix/vector
–  http://techblog.netflix.com/2015/04/introducing-vector-netflixs-on-host.html
•  Netflix Atlas
–  http://techblog.netflix.com/2014/12/introducing-atlas-netflixs-primary.html
•  Heat Maps
–  http://www.brendangregg.com/heatmaps.html
–  http://www.brendangregg.com/HeatMaps/latency.html
•  Flame Graphs
–  http://www.brendangregg.com/flamegraphs.html
–  http://techblog.netflix.com/2014/11/nodejs-in-flames.html
•  Frequency Trails: http://www.brendangregg.com/frequencytrails.html
•  Methodology
–  http://www.brendangregg.com/methodology.html
–  http://www.brendangregg.com/USEmethod/use-linux.html
•  perf-tools: https://github.com/brendangregg/perf-tools
•  eBPF: http://www.brendangregg.com/blog/2015-05-15/ebpf-one-small-step.html
•  Images:
–  horse: Microsoft Powerpoint clip art
–  gauge: https://github.com/thlorenz/d3-gauge
–  eBPF ponycorn: Deirdré Straughan & General Zoi's Pony Creator

Thanks
•  Questions?
•  http://techblog.netflix.com
•  http://slideshare.net/brendangregg
•  http://www.brendangregg.com
•  bgregg@netflix.com
•  @brendangregg
Jun
2015

Monitorama 2015 Netflix Instance Analysis

Recommended

More Related Content

What's hot (20)

Similar to Monitorama 2015 Netflix Instance Analysis (20)

More from Brendan Gregg (20)

Recently uploaded (20)

Monitorama 2015 Netflix Instance Analysis