Presentation is loading. Please wait.

Presentation is loading. Please wait.

Exchange 2010 Project Presentation/Discussion August 12, 2015 Project Team: Mark Dougherty – Design John Ditto – Project Manager Joel Eussen – Project.

Similar presentations


Presentation on theme: "Exchange 2010 Project Presentation/Discussion August 12, 2015 Project Team: Mark Dougherty – Design John Ditto – Project Manager Joel Eussen – Project."— Presentation transcript:

1 Exchange 2010 Project Presentation/Discussion August 12, 2015 Project Team: Mark Dougherty – Design John Ditto – Project Manager Joel Eussen – Project Support Karen McLaughlin – Design

2 Major Goals Exchange 2010 Project

3 Major Goals Bring down statewide total cost of ownership: – Reduce server count – Reduce hardware costs – Reduce administration & support costs – Reduce power & real estate consumption Provide higher availability of email service and data statewide, with a focus on quick recovery Support flexible delegation model Accommodate the data center move with minimal downtime

4 Overview of Exchange 2010

5 Exchange Enterprise Topology Enterprise Network External SMTP servers Mailbox Storage of mailbox items Gateway or Edge Routing & AV/AS Client Access Client connectivity Web services Hub Transport Routing & Policy Web browser Outlook (remote user) Mobile phone Outlook (local user)

6 Enabling Technologies 64 bit platform IO Reduction from Exchange 2003 to 2010

7 IO Reduction: Database Page Size Increased to 32 KB Page 1 Msg Header Page 2 X Page 3 Msg Body Disk Page 4 X Page 5 Msg Body DB Cache Page 1 Msg Header Page 3 Msg Body Page 5 Msg Body 3 Read IO’s Page 1 (32KB) Msg Header, Msg Body Disk DB Cache 1 Read IO Exchange Server 2007 DB Read 20 KB Message Exchange Server 2010 DB Read 20 KB Message 8 KB Pages 32 KB Pages Page 2 (32KB) X Page 1 (32KB) Msg Header, Msg Body

8 Exchange IO Trend +90% Reduction!

9 Reducing cost of ownership

10 IO Reduction: ESE Changes Optimize for new store schema – ESE table for each mailbox (sequential IO), maintain database contiguity – Utilize space efficiently (database compression) Increase database (DB) IO Size – DB page size increased from 8 kilobyte (KB) to 32 KB Page Patching – Self repair databases from DAG Optimized for large mailboxes (+10 GB) & 100K items Optimized for large/slow/low-cost disks (SATA/JBOD)

11 Client Access Server Mailbox Server 1 Mailbox Server 2 Mailbox Server 3 Mailbox Server 6 Mailbox Server 4 AD site: Spokane AD site: Olympia Mailbox Server 5 Exchange 2010 High Availability Overview Failover managed within Exchange Easy to stretch across sites Client Access Server All clients connect via CAS servers Database centric failover

12 Databases Corruption and 1018s Best case scenario: > Move mailboxes to new database. > ECC checksum automatically fixes issues 40% of the time Worst case scenario: > Backups cease > Must take database offline > Eseutil or isinteg to fix > May require restoring and log replay Page patching process works constantly in the background to detect problems. When found, bad pages are repaired automatically, taking information from all valid copies (using an arbitration algorithm).

13 Hard Drives & Servers Failures x HOT SPARES Drive failures covered by RAID on SAN enclosure Recovery Window = Instant Drive failures covered by RAID on DAS enclosure Recovery Window = Instant HOT SPARES x Server Failures require rebuilding the server, restoring Exchange Service, and restoring Exchange data from tape. Recovery window = 8-12 hours. Server failures automatically transfer load to other servers in the DAG. Recovery window = 30 seconds.

14 Site Failures Manual failover must be activated at the Spokane node site. Log files are replayed forward to restore the database up to the point of failure. Recovery Window About 4 hours for service and data restoration. Once site is rebuilt or relocated, all servers must be rebuilt or fixed and backup tapes for all servers must be restored from offsite location. Data restored to the point of most recent backup. Recovery Window 8-40 hours for service restoration Up to 300 hours for data restoration

15 Page 2 (32KB) X High Availability & Quick Recovery Summary Hard drives: Immediate Same Databases: 4-8hrs 30sec Server: 8-12hrs 30secs Site: 2wks 4-8hrs DB Corruption: Unknown Immediate 2003 2010

16 Deleted Item RetentionINBOX DELETED ITEMS DELETED ITEM RETENTION GONE INBOX DELETED ITEMS DELETED ITEM RETENTION GONE Del Or 14 Days 31 Days Later

17 Enhanced Legal Discovery 1.Discoverable mail is stored in multiple places 2.Delegation only on the vault store 3.Cannot guarantee completeness 1.Fewer discoverable sources 2.Delegation on both the mailbox and vault 3.2010 guarantees mailbox completeness

18 Delegation

19 Role Based Access Control (RBAC)

20 Exchange Control Panel (ECP) What is it? A browser based Management client for end users, administrators, and specialists Simplified user experience for common management tasks Accessible directly via URL, OWA & Outlook 14 Deployed as a part of the Client Access Server role RBAC aware

21 Simplify Administration Lower Support Costs Through New User Self- Service Options

22 Reducing Total Cost of Ownership OLYMPIA DATA CENTER SPOKANE NODE SITE Reduce Number of Mailbox Servers Virtualized where Appropriate Consolidate Roles for Support Staff with Delegation and RBAC Full Site Resiliency provides 30 Sec Recovery Window without tape (4 hours in Regional Disaster Scenarios). Hardware Consolidation + Site Consolidation Significant Statewide Energy Savings. Energy inefficient

23 Questions


Download ppt "Exchange 2010 Project Presentation/Discussion August 12, 2015 Project Team: Mark Dougherty – Design John Ditto – Project Manager Joel Eussen – Project."

Similar presentations


Ads by Google