Facebook

Report
TR 1-470
Performance Analysis
for Dummies
Lars Ejskjaer
Greg Ferguson
Agenda
 Understanding Performance Analysis Can Help You Sell
− Gain better understanding of your customers environment
− Grow awareness to possible system bottleneck areas
− Understand changes that can enhance your customers overall
satisfaction
 Getting Started…
 Simple Analysis (1, 2, 3)
 Valuable Tools and Resources
2
The Customer Problem
”My storage is slow…”
3
Your Goal
 Guide the customer to best practices
 Recommend solutions
− Administrative
− Product
 Provide advice in the context of their environment
4
What You Will Need To Know
 Understanding of how the customer’s environment was set up
 Ability to identify missed best practices
− Administrative
− Performance
 Ability to identify common performance issues
5
Getting Started
6
Easily Collect Performance Data - Perfstat
https://communities.netapp.com/docs/DOC-16209
Pro Tip:
Perform multiple smaller iterations versus one larger iteration for better visibility
7
Loading Data into LatX for Analysis
https://latx.netapp.com/
8
Find and View the Perfstat
NetApp Confidential – Limited Use
9
Finding Specific Data
Pro Tip
Use PRESTATS iteration 1 for configuration information
Use POSTSTATS for performance measurements
10
Analysis Strategy
 Understand Configuration
− Aggregates
− Volumes
− Look for configuration errors
 Analyze Performance
− System
− Disk
− Flash
 Make Recommendations
11
The Simple Analysis
Part 1
Disk Configuration
12
Disk Types on the Controller
Prestats - sysconfig -r
Pro Tip
Common disk types today: SAS, BSAS, MSATA
13
RAID Groups
Poststats – statit
Pro Tip
Avoid unbalanced raid-groups in aggregates!
14
32-bit vs 64-bit Aggregates
Prestats – aggr status -v
Pro Tip
 New features work best (require) with 64-bits aggregates
 Using one aggregate type makes the customer operations easier
15
Aggregate Utilization
Poststats - df –A -h
Pro Tip
 An aggregate which is 80% full, should be monitored
 Aggregates over +90% full could impact performance
16
Aggregate Snapshot Copies
Prestats - df –A –h, snap status -A & aggr status –v
Pro Tip
Only SyncMirror and MetroCluster use aggregate snapshots
If not in use:
- Remove schedule
- Release space reservation
NetApp Confidential – Limited Use
17
The Simple Analysis
Part 2
Volumes
18
Volume Space Utilization
Poststats - df & lun stat –v all (or vol status –v)
Pro Tip
Databases like Oracle initialize their data files or data could be static so it could be
fine that the volume is almost full
19
Deduplication
Poststats- stats perfstat_sis
Pro Tip
Savings less than 6-8% are just using resources without a real effect,
unless data is static, turn it off and save resources!
20
Deduplication Runtimes
Poststats - sis status -l
Pro Tip
Deduplication is a low priority process but it still occupies system resources
Work to understand run times to smooth system scheduling for better
performance
21
Misalignment
Poststats - nfsstat –d & lun stat –v all
Pro Tip
Understand what is in these files/LUNs
Logfiles are often ”misaligned”
Verify that virtual machines are aligned
22
The Simple Analysis
Part 3
System Performance
23
What Are Domains In Data ONTAP?
 ONTAP breaks work into groups of processes called domains
 ONTAP schedules work across CPU cores as IT sees best
 This can be seen in Sysstat –M
 Detailed analysis of this is an advanced topic
24
CPU Utilization
Poststats - sysstatM.out
Pro Tip:
Average CPU utilization >70% depicts a very busy controller
CPU Utilization is a generally poor indicator of performance
Look at AVG CPU
– Single CPU (thread) utilization is very informative
25
Writing Data to Disk - CP Type & Time
Poststats – sysstat_1sec.out
Pro Tip
Deferred Back-to-back CP’s are performance killers (type #)
- Data can’t get to disk fast enough
Investigate (ignoring the CPU utilization):
- Mis-alignment
- Disk over commited
Solutions:
- Move load to another controller/aggregate
- Add disks/Flash Cache
26
Aggregate Disk Utilization
Poststats - statit
Pro Tip
Analyze performance impact based on drive utilization (by drive type)
- SATA drives > 50% = busy
- SAS drives > 60% = busy
Statit will give clues about where to move load to
27
Active Volumes
Perfsys Report
Pro Tip
Map volumes to aggregates (aggr status –v)
Identify workloads to move
28
FlashCache
Perfsys Report
Pro Tip
Verify that the system is benefiting from the use of FlashCache
Use PCS and perfstat to verify a customers gains by use of FlashCache
29
Latency in the Environment
Poststats - stats perfstat_cifs/_nfs/_fcp
Pro Tip
High latency = performance impact
Latency requirements vary by applicaton
Analyze the workload and
- Add Disk
- Add FlashCache
- Upgrade the Controllers
30
Review of Findings
 We reviewed
− High disk I/O’s
− Busy volumes
− Disk configuration issues
− Average CPU utilization
− Potential Flash Cache benefits
− Mis-alignments
31
Recommendations
 Resolve mis-alignments
 Consider moving busy volumes
 Add drives and reallocate to even out raid groups
 Add Flash Cache
 Upgrade Controllers
 Add more Controllers within Clustered Data ONTAP
32
Now YOU Can Better Understand System
Performance
Don’t be afraid – performance is no longer such a mystery!
Most importantly, monitor disk utilization!
Have FUN!
33
Important Resources
 The Community site
− http://communities.netapp.com
 ONTAP documentation
(particularly ONTAP Command ref)
− http://support.netapp.com
 Latx – your analysis tool
− http://latx.netapp.com
34
Complimentary Sessions
 TR-1-123 Sizing, Designing, and Presenting
a NetApp Solution
 TR-2-717 Using NetApp Tech Tools to
Create Winning Proposals and Tech Refreshes
 TR-2-315
A Field Guide to Sizing - Part 1
 TR-3-317
A Field Guide to Sizing – Part 2
35
Take an Insight Survey!
NEW!
1) Click on the session number in your agenda.
2) Click on the Surveys Button.
3) Follow the prompts, complete the survey and submit!
Complete this survey by 7PM and be entered
to win one of the following prizes:




1 iPad Mini 16GB Wifi
1 Bose SoundLink Mini Bluetooth
2 Jawbone Up Wristbands (Activity Tracker)
4 NetApp Signature Dry Zone Caps
Went to a different session? Need a translated
survey? Visit the main survey page in the mobile app
to take a daily survey – available in English, Chinese,
Japanese and Korean.
36
Facebook
www.facebook.com/ NetAppInsightAmericas
www.facebook.com/ NetAppInsightEMEA
Twitter
www.twitter.com/ InsightAmericas
www.twitter.com/ InsightEMEA
Tweet friends with #NTAPInsight
37
© 2013 NetApp, Inc. All rights reserved. No portions of this document may be reproduced without prior written consent
of NetApp, Inc. Specifications are subject to change without notice. NetApp, the NetApp logo, and Go further, faster,
are trademarks or registered trademarks of NetApp, Inc. in the United States and/or other countries. All other brands
or products are trademarks or registered trademarks of their respective holders and should be treated as such.
38

similar documents