Optimizing Tableau Server for High Availability

Transcription

1

2 Welcome

3 # T C 1 8 Optimizing Tableau Server for High Availability Kitty Chou Senior Manager, Product Management Tableau / Server and Cloud Platform Dmitry Ryabkov Staff Software Engineer Tableau / Data Management

4 St. Petersburg Russia

5

6 Maryland

7

8 SOS! Tableau Server is down!

9 Objectives How to setup a Highly Available Tableau Server cluster What happens when things go wrong Advanced tips and tricks Disaster Recovery Q&A

10 What is High Availability (HA)? System is available for use with minimum downtime Frequently expressed as a percentage of uptime # of 9 s Availability Downtime/Year Downtime/Month 1 90% 36.5 days 3 days, 1 hour 2 99% 3 days, 15 hours 7 hours, 12 minutes % 8 hours, 45 minutes 43 minutes, 48 seconds % 52 minutes, 34 seconds 4 minutes, 19 seconds Achieved by: 1. Elimination of single points of failure 2. Detection of failures and reliable failover

11 HA in Tableau Server Let s start with a basic Tableau Server installation and look at how we apply these principles Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Application Processes Data Engine File Store Agent Service Manager Client File Service Controller Licensing (New!)

12 HA in Tableau Server Let s start with a basic Tableau Server installation and look at how we apply these principles Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Application Processes Data Engine File Store

13 HA in Tableau Server Must have a minimum of three nodes Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Node 1 Node 2 Node 3

14 HA in Tableau Server Search & Browse File redundancy/replication with multiple Data Engine/ File Store pairs VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Node 1 Node 2 Node 3

15 HA in Tableau Server Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Data Engine File Store Data Engine File Store Node 1 Node 2 Node 3

16 HA in Tableau Server Active/Passive Repository instances across two nodes Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Data Engine File Store Data Engine File Store Node 1 Node 2 Node 3

17 HA in Tableau Server Multiple instances of other server process across nodes Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Data Engine File Store Data Engine File Store Node 1 Node 2 Node 3

18 HA in Tableau Server Multiple instances of other server process across nodes Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Node 1 Node 2 Node 3

19 HA in Tableau Server Multiple instances of server process on a node Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Node 1 Node 2 Node 3

20 HA in Tableau Server Add an external load balancer External Load Balancer Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Search & Browse VizPortal VizQL Server Data Server Cache Server Backgrounder Data Engine File Store Node 1 Node 2 Node 3

21 HA in Tableau Server We re done! External Load Balancer (with the Application Processes) Search & Browse VizPortal Search & Browse VizPortal Search & Browse VizQL Server VizQL Server VizPortal Data Server Data Server VizQL Server Cache Server Cache Server Data Server Backgrounder Backgrounder Cache Server Backgrounder Data Engine Data Engine Data Engine File Store File Store File Store Node 1 Node 2 Node 3

22 HA for

23

24 (Apache Zookeeper) Facilitates coordination amongst processes in a distributed system Quorum is required for consistency. A majority of nodes in the system need to agree on the state Recommended number of nodes: # of nodes # of Zookeeper nodes # of nodes that can fail This is not HA!

25 TSM HA Make TSM Services HA Node 1 Application Process Controller Licensing Client File Service Agent Service Manager Node 2 Application Process Agent Service Manager Node 3 Application Process Agent Service Manager

26 TSM HA Deploy a coordination service ensemble Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Agent Service Manager Node 3 Application Process Agent Service Manager

27 TSM HA Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Agent Service Manager Node 3 Application Process Agent Service Manager

28 TSM HA Add Client File Service to another node in the cluster Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Client File Service Agent Service Manager Node 3 Application Process Agent Service Manager

29 TSM HA TSM Services are now configured for HA Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Client File Service Agent Service Manager Node 3 Application Process Agent Service Manager

30 Repository Failover

31 A single instance of runs on every node of the cluster One of the cluster controller nodes will be selected to be the leader Monitors the health and status of Repository Updates other processes configuration during failover Sends notifications when server processes are unhealthy

32 External Load Balancer Repository Failover When server is healthy, the active repository will continuously stream changes to the passive repository. Cluster Controller Cluster Controller (L) Cluster Controller Active Passive

33 External Load Balancer Repository Failover is also continuously checking on the health of both instances of the repository and storing the state to the ensemble. Cluster Controller Cluster Controller (L) Cluster Controller Active Passive

34 External Load Balancer Repository Failover Node 1 (with the Active Repository) goes down. Cluster Controller Cluster Controller (L) Cluster Controller Active Passive

35 External Load Balancer Repository Failover Since we still have 2 nodes, we still have quorum and system continues to run. s will elect a new leader amongst themselves. Cluster Controller (L) Cluster Controller Passive

36 External Load Balancer Repository Failover After 5 minutes, leader will trigger the failover by updating everyone in the cluster that the active repository is now on Node 2. Cluster Controller (L) Cluster Controller Active

37 External Load Balancer Repository Failover When Node 1 recovers, it is automatically joined back to the cluster and repository will continue to run as the passive. Streaming replications resumes. Cluster Controller (L) Cluster Controller Cluster Controller Passive Active

38 Recovering from Initial Node Failure

39 TSM HA Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Client File Service Agent Service Manager Node 3 Application Process Client File Service Agent Service Manager

40 Recovering from an Initial Node Failure Let s take a look at what happens when Node 1 goes down Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Client File Service Agent Service Manager Node 3 Application Process Client File Service Agent Service Manager

41 Recovering from an Initial Node Failure DON T PANIC! Tableau Server is still running. Try to Repair Node 1. Node 1 Application Process Controller Licensing \ Client File Service Agent Service Manager Node 2 Application Process Client File Service Agent Service Manager Node 3 Application Process Client File Service Agent Service Manager

42 Recovering from an Initial Node Failure 1. Move controller to Node 2 using the Controller recover script (move-tsm-controller) Node 1 Application Process Node 2 Application Process Node 3 Application Process Licensing \ Client File Service 1 Controller Client File Service Client File Service Agent Agent Agent Service Manager Service Manager Service Manager Reference:

43 Recovering from an Initial Node Failure 2. Stop Tableau Server Stop Tableau Server (tsm stop) Node 1 Node 2 Application Process Application Process Node 3 Licensing Application Process \ Client File Service 1 Controller Client File Service Client File Service Agent Agent Agent Service Manager Service Manager Service Manager Reference:

44 Recovering from an Initial Node Failure 3. Move Licensing process from Node 1 to Node 2 & restart Controller Stop Tableau Server (tsm stop) Node 1 Node 2 Application Process Application Process Node 3 Licensing Application Process \ Client File Service 1 Controller Client File Service Client File Service Agent Agent Agent Service Manager Service Manager Service Manager Reference:

45 Recovering from an Initial Node Failure 4. Activate License Key(s) on Node 2 Stop Tableau Server (tsm stop) Activate License Keys on Node 2 Node 1 Node 2 Application Process Application Process Node 3 Licensing Application Process \ Client File Service 1 Controller Client File Service Client File Service Agent Agent Agent Service Manager Service Manager Service Manager Reference:

46 Recovering from an Initial Node Failure 5. Remove Node 1 and all processes (, File Store, ) from topology Node 2 Application Process Stop Tableau Server (tsm stop) Activate License Keys on Node 2 Remove Node 1 from topology Node 3 1 Licensing Controller Client File Service Agent Service Manager Application Process Client File Service Agent Service Manager Reference:

47 Recovering from an Initial Node Failure 6. Add a new node (Node 4) to your cluster and redeploy coordination service ensemble to the new cluster Stop Tableau Server (tsm stop) Add new node to cluster Node 4 Application Process Client File Service \ Agent Service Manager 1 Node 2 Application Process Licensing Controller Client File Service Agent Service Manager Activate License Keys on Node 2 Remove Node 1 from topology Node 3 Application Process Client File Service Agent Service Manager Reference:

48 Recovering from an Initial Node Failure Start Tableau Server (tsm start)! Stop Tableau Server (tsm stop) Add new node to cluster Node 4 Application Process Client File Service \ Agent Service Manager 1 Node 2 Application Process Licensing Controller Client File Service Agent Service Manager Activate License Keys on Node 2 Remove Node 1 from topology Node 3 Application Process Client File Service Agent Service Manager Reference:

49 Advanced Tips & Tricks

50 Tip #1: Use an External Zookeeper Cluster How-to steps: 1. Install a multi-node Tableau Server cluster as usual 2. Stop server 3. Run the Tableau Server installer on 3 additional nodes These will be the dedicated coordination service nodes Reference:

51 Tip #1: Use an External Zookeeper Cluster 4. Using the TSM Web UI to add three additional worker nodes to the cluster with no processes. 5. Open command prompt to run: tsm topology deploy-coordination-service -n node4,node5,node6 tsm status v tsm topology cleanup-coordination-service tsm start

52 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Cache Server Data Engine File Store Data Engine File Store Backgrounder Data Engine

53 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Cache Server Data Engine File Store Data Engine File Store Backgrounder Data Engine Nodes handling Interactive Workloads

54 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Cache Server Data Engine File Store Data Engine File Store Backgrounder Data Engine Background Workloads

55 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Cache Server Data Engine File Store Data Engine File Store Backgrounder Data Engine Repository

56 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Data Engine File Store Cache Server Data Engine File Store Backgrounder Data Engine

57 Tip #2: Optimizing Server by Workload & HA External Load Balancer Search & Browse Search & Browse Backgrounder Search & Browse Search & Browse VizPortal VizPortal Data Engine Repository (A) Repository (P) VizQL Server VizQL Server Data Server Data Server Cache Server Cache Server Data Engine File Store Data Engine File Store Backgrounder Data Engine

58 Tip #3: Deploying to Public Cloud Region Each region is completely independent Availability Zone (AZ) Each AZ is isolated, but the AZs in a region are connected through lowlatency links

59 Tip #3: Public Cloud Deploy to Multiple AZs Deploy to a region with at least 3 AZs Deploy workers into different AZs/subnets Low latency between AZs allow for Tableau Server to run smoothly even though nodes in the cluster are deployed to separate AZs

60 Disaster Recovery

61

62 RPO and RTO Overall Recovery Goals Recovery Point Objective (RPO) Recovery Time Objective (RTO) Time 1 System Operational with Time of last backup Incident data from time point 1

63 Take Regular Backups! (RPO) Make sure you have enough disk space (requires ~2x size of the current files) Move backup files to another location/region or save it to the cloud Secure your backup files

64 Reducing RTO High RTO Low RTO No Standby Cluster (RTO: Several hours/days) Cold Standby Cluster (RTO: A few hours) Dedicated DR Cluster (RTO: Within a few minutes) *DR environments are ideally in a different data center and/or region

65 External Load Balancer Disaster Recovery Cluster Example Primary Cluster (Region A) Node 1 Node 2 Node N Node 1 Node 2 Node N Restore backups from Primary cluster to DR cluster DR Cluster (Region B)

66 Pop Quiz

67 Question 1: What s the minimum number of nodes in your cluster to still be Highly Available?

68 Question 2: What are the two TSM processes that do not have automatic failover/redundancy?

69 Question 3: Where should you store your backups?

70

71 R E L AT E D S E S S I O N S Tableau Server security in depth October 25 2:15pm 3:15pm MCCNO L3 351 Tableau Server on Linux 2018 update October 25 2:15pm 3:15pm MCCNO L3 340

72 Please complete the session survey from the Session Details screen in your TC18 app

73 #TC18 Thank you!

74

75 Section/Chapter Title

76 Section/Chapter Title

77 Section/Chapter Title

78 Agenda Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim

79 Agenda Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim Lorem ipsum dolor sit amet, error possim

80 Slide Title Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here

81 Slide Title (2 Column layout) Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here

82 Slide Title Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here

83 Slide Title (2 Column layout) Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here

84 Slide Title (Image with text) Main topic 1: Subtopic copy goes here Subtopic copy goes here Main topic 2: Subtopic copy goes here Subtopic copy goes here Main topic 3: Subtopic copy goes here Subtopic copy goes here

85

86 Sample Code var pd = require('pretty-data').pd; var xml_pp = pd.xml(data); var xml_min = pd.xmlmin(data [,true]); var json_pp = pd.json(data); var json_min = pd.jsonmin(data); var css_pp = pd.css(data); var css_min = pd.cssmin(data [, true]); var sql_pp = pd.sql(data); var sql_min = pd.sqlmin(data);

87 Sample Code var pd = require('pretty-data').pd; var xml_pp = pd.xml(data); var xml_min = pd.xmlmin(data [,true]); var json_pp = pd.json(data); var json_min = pd.jsonmin(data); var css_pp = pd.css(data); var css_min = pd.cssmin(data [, true]); var sql_pp = pd.sql(data); var sql_min = pd.sqlmin(data);

88 Click to add quote Name Title, Org

89 Click to add quote Name Title, Org

90 Click to add quote Name Title, Org

91 Please complete the session survey from the Session Details screen in your TC18 app

92 Viz Slide Lorem ipsum dolor sit amet, consectetur adipiscing elit. Proin porttitor leo tortor, sed ullamcorper ipsum suscipit in. Curabitur fermentum lacinia lectus non laoreet. Sed volutpat, dui eu rutrum volutpat, nulla mi accumsan dui, non venenatis mauris augue nec lectus.

93 Viz Slide 2

94 Template Apply the template to an existing PowerPoint presentation (that uses a different template) 1. Save this template to your Desktop. 2. Open an existing PowerPoint you wish to update. 3. Click Design, then scroll down and select "Browse for Themes." 4. Browse to the TC17 template file (.potx) you saved to your Desktop and click Open. 5. The template should update the design and font.

95 Theme Colors Text & slide background colors Accent colors Text/ Background Light 1 Text/ Background Light 2 #00AEC7 #AFCC35 #A72D92 F48721 #1B1852 Type and background combinations must meet a minimum 4.5:1 contrast ratio or greater for accessibility and large format screen legibility. Examples above demonstrate correct color use for on-screen applications.

96 Typography Fonts This template has been formatted with Arial (bold) and Arial which are standard fonts set in the Slide Master to avoid compatibility issues. Slide Titles and Headers Slide Title and Header text should be set in Arial (bold) have each word capitalized Main topics and Headers should be set at 32pt when possible, use colored text for visual differentiation/focus. Body Copy Body copy should be set to Arial 24pt when possible. Try to limit each slide to a maximum of 3 font sizes. Type Tips Create visual differentiation/focus by using scale and color versus using bullets.

97 Grid/Guidelines To view the grid and guidelines, check the Guides box in the Show section under the view tab. To further aid in alignment of objects, click the expand icon in the bottom-right corner of the Show section and check Display smart guides when shapes are aligned.

98 Icon set Access Big Data Any Data Anywhere Any Platforms Best Practices Calculations Library Collaborate Community Dashboard Data Blending Deep Statistics Embedded Analytics Folder Location Data Sources Mobile People Settings Search Security Server Admin Tableau Desktop Tableau Server Tableau Public Tableau Online Tableau Prep These icons are provided so you can use them to for diagrams showing architecture, workflow, etc. Icon colors can be modified by right-clicking item and selecting theme color.

99 Tables To modify table, first click anywhere in table, so the Table Tools menu is highlighted at top Layout > Shrink or expand column widths by adjusting the Cell Size, or set them to same size with Distribute To modify the table layout, click Table Tools > Layout To modify the table style, click Table Tools > Design Layout > Use Alignment settings to adjust text alignment and cell margins Layout > To add rows, click into cell and choose, Insert Above or Insert Below Layout > To add columns, click into cell and choose Insert Left or Insert Right Tip: To quickly add a row, place cursor in this last cell and hit Tab key

100 Sample Table (Option A) Header 3 Header 4 Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content

101 Sample Table (Option B) Header 1 Header 2 Header 3 Header 5 Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content Content

102 #TC18 Thank you! Contact or CTA info goes here

103

104

105 S E S S I O N R E P E AT S Session Name Day Time Time Location Session Name Day Time Time Location