Curran- CAPSPUG Presentation

Report
Real World SharePoint 2013 Implementation
Patrick Curran, MCT
6/18/2014
#CAPSPUG
© 2011 PLANET TECHNOLOGIES, INC.
Planet Technologies








Founded in 1998
Microsoft Partner with Five Gold Competencies, 11 Silver Competencies
One of only 35 Microsoft Nationally Managed Partners
Six Time Microsoft Federal Partner of the Year and Three Time Microsoft SLG Partner of the Year
Highly Skilled Staff
– Microsoft Certified Masters
– Microsoft MVP’s (SharePoint, Exchange)
– Microsoft Certified Trainers
– Microsoft Certified Partner for Learning Solutions (CPLS)
– 130+ Microsoft Consultants
Office Locations – DC, Denver, Harrisburg, Redmond, Europe, Middle East and Africa (EMEA)
Vertical Markets include Federal, State and Local Government, Education, Healthcare and Private
Sector
Work Experience in 42 States and 33 Countries
© 2011 PLANET TECHNOLOGIES, INC.
Patrick Curran
 MCT, MCPD, MCITP, MCP, MCTS, MCSA
 Director, Federal Practice, Planet Technologies.
 SharePoint Architect.
 Co-Author of Microsoft SharePoint 2013: Designing and
Architecting Solutions.
 Working with SharePoint since 2003.
PCfromDC.com
@PCfromDC
© 2011 PLANET TECHNOLOGIES, INC.
Microsoft® SharePoint® 2013: Designing and Architecting Solutions
© 2011 PLANET TECHNOLOGIES, INC.
Agenda
1. Introduction to ContosoPharma.
2. Introduction to ContosoPharma IT.
3. Review architecture and verify solution.
4. Review supporting documentation.
5. Review implementation of Server 2012 and SQL2012.
6. Review implementation of WAC (OWA) farm.
7. Review implementation of Server 2012 and SP2013.
8. Review implementation of Workflow Manager farm.
9. Preparing for Disaster Recovery
10. Review a Different Production Environment
Meet ContosoPharma
We have won a bid for deploying SharePoint 2013 at Contoso
Pharmaceuticals. The winning proposal is to deploy a
SharePoint 2013 farm for 25 people growing to 100 people.
IT has already purchased 2 servers!
Let’s get started!
© 2011 PLANET TECHNOLOGIES, INC.
Meet ContosoPharma
Make sure that if you use acronyms that everyone in the room
understands what you mean. Do NOT assume that even the
most basic of concepts will be understood. For example, HA,
means High Availability, meaning if a server or service fails that
there is a continuity of operation with minimal end-user impact…
While using network load-balancing will allow for distributed load
on the network, it is needed for high availability, even if the farm
is only for 50 people.
© 2011 PLANET TECHNOLOGIES, INC.
Kickoff Meeting Scenario
Day 1 Conclusion:
Client would like 1 SharePoint farm, created in HA (high
availability mode). Already purchased hardware (although
needed specs still).
Schedule meeting with IT Staff to validate requirements without
CIO in the room.
© 2011 PLANET TECHNOLOGIES, INC.
Kickoff Meeting Scenario
Farm drawing at this point
© 2011 PLANET TECHNOLOGIES, INC.
SQL Server Always-On
http://technet.microsoft.com/en-us/library/hh913923(v=office.14).aspx
© 2011 PLANET TECHNOLOGIES, INC.
SQL Server Always-On
© 2011 PLANET TECHNOLOGIES, INC.
New for SharePoint 2013!
Office Web Application
(WAC- Web Access Companion) Server
• New “disconnected” Service
• Cannot be installed on a SharePoint Server
© 2011 PLANET TECHNOLOGIES, INC.
New for SharePoint 2013!
Workflow Manager 1.0
(WAWS- Windows Azure Workflow Server)
•
•
•
•
Another “disconnected” Service
Can be installed on a SharePoint Server
1 or 3 servers (not 2 or 4). Yes, only 1 or 3
A great example of the new SharePoint application model. Workflow
communications happen via the core app model technologies, CSOM and REST,
and authenticates through OAuth.
Current Versions
• Workflow Manager = 1.0 Refresh
• Service Bus = 1.1
© 2011 PLANET TECHNOLOGIES, INC.
IT Meeting Scenario


This is basically the meeting to really see what is going on
within the organization and to see what they really want /
need. Also verify with IT any questions that were left
outstanding and see if IT and the CIO are on the same page.
Make sure that if you use acronyms that everyone in the
room understands what you mean. Do NOT assume that
even the most basic of concepts will be understood. For
example, DR, means Disaster Recovery, meaning if your
primary location fails to exist that there is a continuity of
operation without any or minimal end-user impact…
© 2011 PLANET TECHNOLOGIES, INC.
I am not referring to Dr Hibbert or Dr Nick when I say DR
IT Meeting Scenario




IT Hands over Pre-Purchased server documentation
IT has been informed that a Dev environment is a good idea
And a 3rd server has magically been allocated to the project.
Start conversation of SLAs, RTOs, and RPOs
Server Hardware Dell PowerEdge R720 (x3)
CPUs
Cores / CPU
Threads / CPU
Total Threads
RAM
Drive Letter
Drive Function
RAID Level
C:
Host OS & VM SQL Temp
Files
1 (Mirror)
Individual Drive Size 146GB
Spindle Speed
15k
Total Drive Size
146GB
© 2011 PLANET TECHNOLOGIES, INC.
2
8
16
32
64 GB
D:
E:
VM SQL Data
VM OS Drives
Files
10 (Mirror/Stripe) 10 (Mirror/Stripe)
F:
VM SQL Data
Files
10 (Mirror/Stripe)
G:
VM SQL Backup
Files
NA
H:
Host "Hot
Spare"
NA
Z:
Host DVD
Drive
NA
500GB
15k
1000GB
900GB
10k
1800GB
900GB
10k
900GB
900GB
10k
900GB
NA
NA
NA
900GB
10k
1800GB
IT Meeting Scenario

http://technet.microsoft.com/en-us/library/cc262485(v=office.15).aspx#hwLocServers

SharePoint Server Minimums (3 tier)
– CPU: 4 cores
– RAM: 12 GB
SQL Server Minimums (small deployment < 1,000 users)
– CPU: 4 cores
– RAM: 8 GB

Minimums Host A
© 2011 PLANET TECHNOLOGIES, INC.
Host B
CPU
24 threads 20 threads
RAM
68 GB
56 GB
IT Meeting Scenario
Oh, so you don’t want own any datacenter CALs
and you don’t want to purchase any?

http://www.microsoft.com/en-us/server-cloud/products/windows-server-2012r2/buy.aspx#fbid=-RPCoaYVf-4


Server 2012R2 Datacenter Processor + CAL $6,155
Server 2012R2 Standard Processor + CAL $882
© 2011 PLANET TECHNOLOGIES, INC.
IT Meeting Scenario
Host
OS
C:\
Data Drives
D:\
E:\
Backups
F:\
Hot
Spare
Z:\
Temp
Drives
NIC Team
Contoso Internal
Network
© 2011 PLANET TECHNOLOGIES, INC.
G H
VM
OSs
W ork fl ow
Network
Log Drives
Host
M achine
Ne twork
IT Meeting Scenario
Farm drawing at this point
Production
Production
Cloud1
Development
Cloud2
1
3
Cloud3
5
staff.contosopharma.com
wf.contosopharma.com
4
2
wac.contosopharma.com
dev-staff.contosopharma.com
dev-wf.contosopharma.com
6
dev-wac.contosopharma.com
Office
Workflow
Web
Manager
Applications
© 2011 PLANET TECHNOLOGIES, INC.
SQL2012
SP2013
Oh, So That’s Not What You Meant?
After updated drawings are handed back over to IT for review, it is
determined by the client that they want one production farm in DC
and another COOP farm in VA for disaster recover.
COOP = Continuity of Operations
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
As of April 2014, it is now supported to do an
asynchronous commit of content and services
databases. For a complete listing of all the databases
supported review:
http://technet.microsoft.com/en-us/library/jj841106(v=office.15).aspx
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
Farm drawing at this point
Production
Development
Cloud1
1
staff.contosopharma.com
wf.contosopharma.com
2
COOP
Cloud3
5
Cloud2
3
dev-staff.contosopharma.com
dev-wf.contosopharma.com
staff.contosopharma.com
wf.contosopharma.com
4
6
wac.contosopharma.com
© 2011 PLANET TECHNOLOGIES, INC.
dev-wac.contosopharma.com
wac.contosopharma.com
Oh, So That’s Not What You Meant?
What you really want is a stretched farm…
Stretched Farm = “A single farm that has servers in multiple physical locations”
• Originally not supported for SharePoint 2013
• Must be less than 1 millisecond latency between all the SQL Servers and the
front-end Web servers in one direction, and at least 1 gigabit per second
bandwidth.
http://blogs.msdn.com/b/zwsong/archive/2011/04/04/killing-two-birds-with-onestone-sharepoint-ha-and-dr-with-stretch-farm-and-everything-you-want-to-knowabout-it.aspx
• There must be a highly consistent intra-farm latency of <1ms, 99.9% of the time
over a period of ten minutes. (Intra-farm latency is commonly defined as the
latency between the web front-end and database servers)
http://blogs.technet.com/b/stefan_gossner/archive/2013/04/04/update-on-stretchfarm-support-in-sharepoint-2013.aspx
• Distributed topologies across one or more geographic areas is not supported.
(huh?)
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
What you really want is a stretched farm…
What does this really mean?
• I cannot find another reference other than Stefan’s that says no geographic border
crossing of data.
• Latency is latency… <1ms does not care if data is traveling 1m or 1km
• Reality is that a bad switch setting or a slow network will not even allow for a
supported SharePoint farm let alone a stretched farm.
Sorry Contoso
Pharmaceuticals, not
supported, not doing it…
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
What you really want is a stretched farm…
Dark Fibre: unlit fibre, meaning that there are no other lasers using that piece of glass
• Typically set as a 10GB/s link requiring 2 fibre channels 1 for TX and 1 for RX.
• Fibre bandwidth does not change latency
• You cannot defy physics, the speed of light in fibre is 300 x 10^6 m/s * 0.66 = 200 x
10^6 m/s (roughly 200km / ms or 124 miles / ms)
• If out of fibre channels, you can always increase bandwidth by multiplexing
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
What you really want is a stretched farm…
If you are one of those, “I hate physics, can you please just tell me what I need to
know” people:
Copper
Fibre
Formula for distance in
0.1ms
18000 / (180 x 10^6 m/s)
200000 / (200 x 10^6 m/s)
Distance in 1ms
18km (11 miles)
200km (124 miles)
© 2011 PLANET TECHNOLOGIES, INC.
Oh, So That’s Not What You Meant?
Farm drawing at this point
Development
Production-HQ
Cloud3
5
Production-COOP
Cloud1
Cloud2
1
3
staff.contosopharma.com
wf.contosopharma.com
dev-staff.contosopharma.com
dev-wf.contosopharma.com
6
dev-wac.contosopharma.com
© 2011 PLANET TECHNOLOGIES, INC.
4
2
wac.contosopharma.com
Review Current Documentation
 Note: At this point in time, HQ and COOP sites are in different
IP ranges.
HQ
COOP
172.20.0.xxx
172.20.8.xxx
Issues
 Requires a lot of IP addresses
 DNS replication and DHCP updating needs to happen quickly
(was not happening with ContosoPharma)
© 2011 PLANET TECHNOLOGIES, INC.
Review Current Documentation
Farm drawing at this point… What’s Wrong?
Development
Production-HQ
Cloud3
5
Production-COOP
Cloud1
Cloud2
1
3
staff.contosopharma.com
wf.contosopharma.com
dev-staff.contosopharma.com
dev-wf.contosopharma.com
6
dev-wac.contosopharma.com
© 2011 PLANET TECHNOLOGIES, INC.
4
2
wac.contosopharma.com
Review Current Documentation
Farm drawing… Final?
Cloud3
Cloud1
5
1
wf.contosopharma.com
Cloud2
staff.contosopharma.com
wf.contosopharma.com
3
ca.contosopharma.com
dev-staff.contosopharma.com
2
6
4
wac.contosopharma.com
dev-wac.contosopharma.com
Office
Workflow
Web
Manager
Applications
SQL2012
© 2011 PLANET TECHNOLOGIES, INC.
Contoso Internal Network NIC Team
Contoso Virtual NIC
SP2013
SQL Server Web Server
Contoso Workflow NIC
Server Management NIC
Contoso Internal Network
Server, SQL, and WAC Installation
 Have scripts prepared for setting up the VMs on the host
server.
– You should have different scripts for SQL and SharePoint
 Have scripts prepared to set up the virtual server from a
“SYSPREP’d” state.
 As of April 2014 CU, SharePoint now has SQL Availability
Group cmdlets.
–
–
–
–
Get-AvailabilityGroupStatus
Add-DatabaseToAvailabilityGroup
Remove-DatabaseFromAvailabilityGroup
Trevor Seward has a nice post
•
http://thesharepointfarm.com/2014/05/sharepoint-database-availability-group-cmdlets/
© 2011 PLANET TECHNOLOGIES, INC.
Server, SQL, and WAC Installation
 Review database management for SharePoint 2013
– http://technet.microsoft.com/en-us/library/cc825329(v=office.15).aspx
– Move site collections between databases
– Move all databases
 DO NOT CREATE A NLB CLUSTER ON THE SQL SERVER
ALWAYS-ON SERVERS!
© 2011 PLANET TECHNOLOGIES, INC.
Server, SQL, and WAC Installation
 Create your SQL backup jobs now
– If you have not already started your backup and restore strategy with client,
start it now.
– Do not put all of your eggs in one basket
– Content is KING!
© 2011 PLANET TECHNOLOGIES, INC.
Server, SQL, and WAC Installation
 Anything WAC related refer to Wictor Wilén
– http://www.wictorwilen.se/spc14-scripts-for-mastering-office-web-apps-2013operations-and-deployments
 Configuring Office Web Apps in SharePoint 2013 (Steve Peschka)
– http://blogs.technet.com/b/speschka/archive/2012/07/23/configuring-officeweb-apps-in-sharepoint-2013.aspx
 Configure Office Web Apps for SharePoint 2013 (TechNet)
– http://technet.microsoft.com/en-us/library/ff431687.aspx
 Office Web Apps 2013: Patching your WAC farm with no downtime
– http://www.wictorwilen.se/office-web-apps-2013-patching-your-wac-farm-withno-downtime
 Apply software updates to Office Web Apps Server
–
http://technet.microsoft.com/en-us/library/jj966220.aspx
© 2011 PLANET TECHNOLOGIES, INC.
Server, SP2013, and WAWS Installation




Enable all required features on the server.
Install all SharePoint and Workflow bits.
SYSPREP the server.
Export the image from Hyper-V and validate.
–
–
Create new VM run through scripts and make sure everything works.
When new bits are updated in farm, update bits on image, sysprep again retest.
 Create all SharePoint servers in the farm based off of this image.
 Have scripts prepared to set up the virtual machine from a “SYSPREP’d”
state. Including having the scripts available to add the new server to the
SharePoint Farm.
© 2011 PLANET TECHNOLOGIES, INC.
Server, SP2013, and WAWS Installation
 Name your workflow instance appropriately, don’t just go with the default
name of “SharePoint”
$site = "https://staff.contospharma.com"
$workflowHostUri = "https://wf.contosopharma.com:12290"
$scopeName = “SharePoint-Production"
Register-SPWorkflowService -SPSite $site `
-WorkflowHostUri $workflowHostUri `
-ScopeName $scopeName
© 2011 PLANET TECHNOLOGIES, INC.
Server, SP2013, and WAWS Installation
 Anything WAWS related refer to Spence Harbar
–
–
–
–
–
Has an excellent 4 part blog post on Workflow
http://www.harbar.net/articles/wfm1.aspx
http://www.harbar.net/articles/wfm2.aspx
http://www.harbar.net/articles/wfm3.aspx
http://www.harbar.net/articles/wfm4.aspx
 Just in case you want not as good information, here is msdn
–
http://msdn.microsoft.com/en-us/library/office/jj163276.aspx
 And here is TechNet
–
http://technet.microsoft.com/en-us/library/jj658588(v=office.15).aspx
© 2011 PLANET TECHNOLOGIES, INC.
AppFabric (Distributed Cache Service) Farm
© 2011 PLANET TECHNOLOGIES, INC.
AppFabric (Distributed Cache Service) Farm
 By default (and in SharePoint), cached items in an
AppFabric Named Cache are stored only once
across the entire cluster.
– SharePoint uses 10 Named Cache Stores
Cache Name
TTL
Configuration Location
ActivityFeed
168 hours
UserProfileApplication.Feed
CacheTTLHours
ActivityFeedLMT
168 hours
UserProfileApplication.
FeedCacheLastModifiedTi
meTtlDeltaHours
LogonToken
10 hours
SPSecurityTokenServiceMa
nager.WindowsTokenLifeti
me
ServerToAppServerAccess
Token
24 hours
(hard coded)
ViewState
31 minutes
SPWebApplication.FormDig
estSettings.Timeout + 1
Access
1 hour
(hard coded)
Bouncer
1 hour
(hard coded)
Search
SecurityTrimming
Default
© 2011 PLANET TECHNOLOGIES, INC.
AppFabric (Distributed Cache Service) Farm
 If the cache host storing the cached item crashes or is shutdown “nongracefully”, that item is no longer available in the cache.
 AppFabric is NOT SharePoint’s responsibility
 Currently at v1.0 CU5
–
http://www.wictorwilen.se/how-to-patch-the-distributed-cache-in-sharepoint-2013
 Must Reads
–
–
–
http://blogs.msdn.com/b/besidethepoint/archive/2013/03/27/appfabric-caching-and-sharepoint-1.aspx
http://blogs.msdn.com/b/besidethepoint/archive/2013/03/27/appfabric-caching-and-sharepoint-2.aspx
http://blogs.technet.com/b/uktechnet/archive/2013/05/07/guest-post-distributed-cache-service-insharepoint-2013.aspx
© 2011 PLANET TECHNOLOGIES, INC.
Preparing for Disaster Recover
Haven’t you already done this?
You have:
• SQL Always-On up and running moving data asynchronously up to Azure (or DR
location of choice) to warm standby farm.
• Standby images to add servers to Farm if need be.
• Scripts have been tested and proven.
• Documentation created on what to do in case of an emergency.
• Not just tested once and forgotten, but added to and updated after each test
• Save master file in your SharePoint Farm, but DO NOT save current
version in SharePoint. Have physical copies printed, .pdf’s emailed, just
something to walk your team through a restore if SharePoint goes down.
© 2011 PLANET TECHNOLOGIES, INC.
Another Type of Production Farm
Cloud1
Cloud3
1
Cloud2
9
16
10
17
11
18
12
19
13
20
staff.vandapharma.com
2
3
5
6
owa.vandapharma.com
wfm.vandapharma.com
Content
14
7
21
Services
15
8
Workflow
© 2011 PLANET TECHNOLOGIES, INC.
22
Questions?
Email: [email protected]
Twitter: @PCfromDC
Blog: http://PCfromDC.com
Thank You!!!
Email: [email protected]
Blog: PCfromDC.com
Twitter: @PCfromDC
© 2011 PLANET TECHNOLOGIES, INC.

similar documents