SlideShare a Scribd company logo
1 of 8
Download to read offline
Is BranchCache right for remote,
serverless software distribution?

1E Technical Whitepaper
Microsoft BranchCache and System Center
Configuration Manager 2007


Abstract
BranchCache is a new feature available in Windows Server 2008 R2 and Windows 7 that
reduces WAN bandwidth usage and improves application responsiveness when workstations
in a remote location access content from the head office or data center by downloading and
caching content on the local network as it is requested, making it immediately available to
other clients that subsequently request the same content.

This paper examines the BrachCache functionality specifically in the context of software
distribution using System Center Configuration Manager 2007 to determine if it is an optimal
solution for the deployment of software, patches and operating systems to remote,
serverless branches.

Key Findings

       BranchCache requires content to be hosted on Windows Server 2008 R2 and
        participating clients to have Windows 7 or Windows Vista with BITS 4.0 installed
       BranchCache does not require a local server when operated in Distributed Cache
        mode, which is ideal for environments of around 50 clients. Microsoft recommend
        the Hosted Cache mode (which requires a local Windows Server 2008 R2 server in
        the branch) for remote locations that include laptops
       BranchCache is a very efficient method for distributing „ad-hoc‟ content, including
        software distributions and updates that are initiated by the user at an arbitrary time
        (i.e. non-mandatory Advertisements that do not have a scheduled install time).
       When the same content is requested simultaneously by several clients (as is the
        case with scheduled software and patch deployments), some of the content may be
        downloaded across the WAN by several clients. If the content is HTTP, BITS is used
        which may result in link saturation when multiple clients simultaneously download
Is BranchCache right for Software Distribution?

Copyright © 1E LTD 2011

Version 1.0 document revision 1

Author: Simon Burbidge & Dave Fuller

All rights reserved. No part of this document or of the software (“the software”) to which it
relates shall be reproduced, adapted, stored in a retrieval system, or transmitted by any means,
electronic, mechanical, photocopying, recording, or otherwise, without permission from 1E Ltd.
It is the responsibility of the user to use the software in accordance with this document and 1E
Ltd shall not be responsible if the user fails to do so. Although every precaution has been taken
in the preparation of this document, 1E Ltd and the authors assume no responsibility for errors
or omissions, nor shall they be liable for damages resulting from any information in it.

Trademarks

1E is a trademark of 1E Ltd registered in the UK, with registration applied for in the US and EU.
The 1E device is a trademark of 1E Ltd with registration applied for in the UK, US, EU and
Australia. NIGHTWATCHMAN is a trademark of 1E Ltd registered in the US, with registration
applied for in the EU and Australia. 1E WAKEUP is a trademark of 1E Ltd. MICROSOFT,
WINDOWS SERVER 2008, WINDOWS VISTA, WINDOWS XP, WINDOWS 2000 and
BRANCHCACHE are all trademarks of Microsoft Corporation in the United States and other
countries.
Contents                Page i



Contents
Introduction 1

How does BranchCache work?                                                                                               1

BranchCache Modes .......................................................................................................1

Discovery and Download .................................................................................................1

Caching 2

Setup and configuration ..................................................................................................2

The Remote Branch Software Distribution Challenge                                                                        2
How does BranchCache measure up?                                                                                         3

No requirement for a local server .....................................................................................3

Efficient use of available bandwidth ..................................................................................3

No dependency for any one client to be available at all times ...............................................4

Efficient use of available disk capacity on local workstations .................................................4

Conclusion        5




Is BranchCache right for remote, serverless software distribution?
Introduction      Page 1




Introduction
BranchCache is a new feature of Microsoft Windows Server 2008 R2 and Windows 7 (also
available on Windows Vista with BITS 4.0 installed). It provides a software-based WAN
optimization solution for branch environments enabling remote content requested by clients in
the branch to be cached locally, either on a dedicated server or among peer workstations, so
that it is available locally to other clients in the branch that may require the same content later
rather than having them download it again over the WAN.

Once set up, BranchCache is transparent to the application and is used to manage HTTP, HTTPS
and SMB2 traffic, so any applications that use these protocols can benefit from BranchCache.
Microsoft included BranchCache support in the SP2 release of System Center Configuration
Manager (ConfigMgr) 2007, enabling its software distribution, software updates and image
deployment features to take advantage of this feature in environments where Windows 2008 R2
and Windows 7 have been deployed and the BranchCache feature configured.

This paper takes a look at the BranchCache feature in detail, focusing on how the feature
performs in remote software distribution, patch management and image deployment scenarios.
Although BranchCache can be configured to use a local Windows Server 2008 R2 server to store
the cache (explained in BranchCache Modes), this paper focuses on the server-less Distributed
Mode, on the assumption that if a local server were available, it would make sense to host a
ConfigMgr Secondary Site or Distribution Point on that server anyway.

How does BranchCache work?
BranchCache Modes
BranchCache can operate in one of two modes: Hosted Mode or Distributed Mode.

Hosted Mode requires a local Windows Server 2008 R2 server in the branch and clients to be
configured with the FQDN of this server. When a client requests content from a remote server
that is not already present on the local BranchCache server, the client downloads the content
but then it is moved to the local server (the content is actually advertised to the server and the
server downloads it from the client). The local cache is maintained on the server and
subsequent requests for the same content from other clients in the branch can be served from
this cache.

Distributed Mode does not require a local server. Instead, content is cached on the workstation
clients and is shared with peer clients when they subsequently require the same content
(remember only Windows 7 or Windows Vista with BITS 4.0 clients can participate). Distributed
Mode is suitable for smaller environments (Microsoft recommends up to 50 clients). Clients
participating in peer-to-peer content transfer must be in the same subnet.

Microsoft recommends using Hosted Mode for branches with more than 50 client and branches
that have laptops that may be removed from the network.

Discovery and Download
BranchCache is transparent to the application. When the application running on a BranchCache
enabled client requests content from a BranchCache enabled server over HTTP/HTTPS or SMB2,
the BranchCache client will request content metadata from the remote server. Metadata is
generated on the source server each time it is requested by the client and consists of block
hashes, segment hashes and a segment key, a segment being a collection of blocks. Typically
the hashing results in a compression ratio of 2000:1, meaning that the metadata is typically
1/2000 (or 0.05%) of the size of the source data. A BranchCache client will always request this
metadata from the source server regardless of whether the data already exists locally to ensure
that the source content has not changed.

Once the metadata has been downloaded, in Distributed Mode clients perform a local discovery
(using the Branch Cache Discovery Protocol, which in turn uses the WS-Discovery protocol) to
determine if any of the segments are available within the branch before resorting to
downloading these from the remote server. HTTP/HTTPS content is downloaded using
Background Intelligent Transfer Service (BITS).

Interestingly, when using SMB2, BranchCache can be configured to be invoked according to
connection latency (a threshold set using NETSH) so allowing BranchCache to only be used


Is BranchCache right for remote, serverless software distribution?
The Remote Branch Software Distribution Challenge            Page 2



when the connection to the server is performing poorly. In contrast, BranchCache using BITS
does not possess this intelligence and is always invoked if enabled.

Note that BranchCache does not download files smaller than 64KB (the requesting application
will receive these through normal SMB2 / HTTP transfer).

Caching
When BranchCache is implemented in Distributed Mode, each client is allocated a proportion of
the total (not free) disk size. By default this is 5% but can be adjusted using NETSH. When the
BranchCache service first starts, the cache is initialized with sparse files (0 data files with a .PDS
extension). These will eventually host the downloaded package and hash data. During testing it
was seen that when BranchCache is actually used for the first time, it appears to instantly
reserve half its cache allocation with the 0 byte data .PDS file being expanded to 50% allocated
cash size (regardless of the size of the actual package being downloaded).

It should be noted that the BranchCache is completely independent from the ConfigMgr client
cache. The ConfigMgr client will populate its own cache from wherever BranchCache gets the
content from. Using BranchCache with ConfigMgr therefore requires sufficient disk space on the
client to host the content in both the BranchCache and the ConfigMgr cache.

Setup and configuration
As long as your source servers are running Windows Server 2008 R2 and your workstations are
on Windows 7 (or Vista with BITS 4.0), setting BranchCache up is relatively straightforward.
BranchCache is installed as a Feature in the Server Manager interface. If you want to support
file transfer (SMB2) with BranchCache you need to add the BranchCache for network files role in
the File Services server role (also managed through Server Manager).

On clients, BranchCache is disabled by default. Most organisations would use Group Policy to
enable and configure BranchCache, typically defining the cache mode, size and location.
BranchCache can also be enabled and managed using the NETSH command-line utility. The
NETSH command-line tool can also be used to manage the local cache on a client (such as
flushing it) and can be used to display configuration settings and cache content.

The Remote Branch Software Distribution Challenge
Remote branch offices will typically range from a handful of clients to perhaps a couple of
hundred. Often these will be a healthy mix of desktops and laptops, with some laptops being off
the network for periods of the time. In many cases there is no facility (or desire) to host servers
in these locations, and often network bandwidth to the head offices and datacenters is highly
contested.

The distribution of software (applications, patches and OS images) in these locations is
somewhat different to the user-initiated applications (e.g. accessing documents from a
SharePoint server) commonly used to demonstrate BranchCache functionality.

Many software and patch deployments require content to be distributed to all machines in the
branch at the same time (typically the case with Software Updates, but also with the planned
roll-out of a new application). Software applications and updates may be distributed in the form
of a single „self-extracting‟ file (e.g. in MSI or EXE format), or may involve the distribution of
several files. Distribution of OS images requires the transfer of very large WIM files and the
process of deploying the image to a workstation will typically require additional package content
(applications) not included in the image.

These deployments have to contend with other applications using the WAN and must avoid
adversely impacting the normal flow of business relying upon it. In order to achieve deployment
within an acceptable time period, WAN traffic should be minimised and bandwidth used
efficiently. This involves ensuring the content only traverses the WAN once and is shared out
within the branch (rather than each client obtaining it from the source server), and ensuring the
download process is aware of existing traffic levels and is able to throttle demand on the link
accordingly, taking advantage of „quiet‟ periods but backing off when other traffic requires the
bandwidth.

In a serverless branch, the dependency is on the client workstations to provide the local cache.
It is therefore important that the workstation that hosts the cache is available at all times, or
that the cache is replicated on several machines to provide resilience.




Is BranchCache right for remote, serverless software distribution?
How does BranchCache measure up?          Page 3



The ideal solution for software distribution to remote branches would have the following
attributes

       No requirement for a local server
       Efficient use of available bandwidth, making use of bandwidth as it becomes available
        but ensuring it is never saturated and ensuring content is only transferred to the branch
        once
       No dependency for any one client to be available at all times
       Efficient use of available disk capacity on local workstations

How does BranchCache measure up?
1E conducted research into the use of BranchCache in the software distribution scenario,
specifically with the use of System Center Configuration Manager (ConfigMgr) 2007 Service Pack
2, which introduced specific integration with BranchCache. This section summarises the results
of our testing in relation to the ideal attributes listed above.

No requirement for a local server
Implementing BranchCache in Distributed mode supports this requirement. Windows 7 and
Windows Vista clients with BITS v4.0 can participate in the peer-caching. Microsoft recommends
that Distributed mode be used for branches up to 50 clients. If clients span multiple subnets
then Distributed mode results in one download per subnet.

In the Key BranchCache Design Points section of the Microsoft Windows Server 20087 R2
BranchCache Design Guide 1 it states that the BranchCache data can only be decrypted by
authenticated clients that are members of the same domain as the content server, although we
have not verified this requirement.

Efficient use of available bandwidth
When multiple clients request the same content simultaneously, as is the case when a
deployment is scheduled at a specific time, we observed that all clients begin to download data
from the source server simultaneously. (The BranchCache performance counters were used to
measure the volume of data downloaded from the WAN, downloaded from a peer cache and
served to others on the network).

This is expected as every client must obtain the metadata from the source to ensure data
integrity, and as previously noted this should amount to 0.05% of the content size if 2000:1
compression is achieved. However, we observed that overall, the amount of data downloaded
across the WAN to the branch was between 2 and 11% greater than the content size (i.e. the
overhead was between 2 and 11%), suggesting that clients were simultaneously downloading
content as well as metadata. The tables below show some examples of our test results.

       Bytes from server is the amount of data transferred across the WAN from the remote
        server
       Bytes from Cache is the amount of data that the client obtained from the cache on
        another local workstation
       Bytes Served is the amount of data that each client provided to other clients from its
        own cache

5MB Single file package deployment
        Client       Bytes from             Bytes from         Bytes Served
                     Server                 Cache
        A            204KB                  4.8MB              1.6MB
        B            347KB                  4.6MB              196KB
        C            5MB                    0                  7.6MB

In this test, all clients began downloading from the server simultaneously. After about 20
seconds, A and B switch to peer caching. The bytes served figures indicates that A and B cached


1
 http://www.microsoft.com/downloads/details.aspx?FamilyID=CBF75A21-BC09-4824-B128-
1A24EE71A9AA&displayLang=en



Is BranchCache right for remote, serverless software distribution?
How does BranchCache measure up?              Page 4



from C as well as from each other. As noted earlier, the file metadata (data that must be first
downloaded from the server) should be using a compression ratio of 2000:1 (0.05% of file size)
or 2.5KB for the test example above. The results show that clients downloaded substantially
more than this amount from the server, suggesting that either the desired compression ratio
was not achieved, or clients were downloading content as well as metadata from the server,
which is inefficient.

15MB package containing 10 unique files
        Client         Bytes from            Bytes from            Bytes Served
                       Server                Cache
        A              5.1MB                 7.9MB                 9.9MB
        B              8MB                   6MB                   13.5MB
        C              2.2MB                 12MB                  2.5MB

In this scenario it is obvious that the clients are simultaneously downloading significant amounts
of content directly from the server. As these clients are using BITS to download the content, it is
not possible to dynamically make optimum use of the available WAN bandwidth 2, and as more
clients participate in the download, the chances of saturating the WAN link are increased.

5MB 10 file package (all files had identical content)
        Client         Bytes from            Bytes from            Bytes Served
                       Server                Cache
        A              4KB                   5MB                   0
        B              500KB                 3.8MB                 458KB
        C              570KB                 3.2MB                 500KB

The figures illustrate an interesting feature of BranchCache. Although we have a 5MB package,
we have only downloaded a total of approximately 1MB from the server. The Program ran
successfully with the entire package present within the ConfigMgr client cache. BranchCache
was able to recognise the fact that the all files shared the same content and effectively reused
data when supplying the clients. BranchCache data discovery behaviour further illustrates this
point. Within this scenario, despite there being 10 files in the source, clients issued only 2
discovery requests as opposed when dealing with unique files, when a discovery request
appears to be issued per file.

No dependency for any one client to be available at all times
The BranchCache Distributed cache model, it was observed that not all participating clients do
not cache the entire content. During a test deployment of a 2.7GB OS image to 3 computers,
only one client had the full 2.7GB in the BranchCache (although the ConfigMgr client cache had
the whole content) while the others had around 1.3GB. The client that had the full image in its
BranchCache was then switched off and the image deployed to another client in the subnet. This
client had to obtain the balance of the image (1.4GB) from the server.

The other tests described in the previous section also revealed that clients had different volumes
of data in their cache, indicating that the caching is indeed distributed and that in order for
other clients to obtain all the content during a subsequent deployment, multiple clients would
need to be available on the network to enable the client to obtain the total content from the
local network.

Efficient use of available disk capacity on local workstations
Although hard disk capacity is pretty cheap these days, using workstation-class hardware to
store distributed content should still make efficient use of the available space, especially if the
workstation is being used to perform a user‟s day job as well.



2
  The BITS client makes use of bandwidth available on the local network adapter and is not aware of
the end-to-end bandwidth available on the WAN. As the local network adapter typically has
significantly more bandwidth (e.g. 100Mb/s) than the WAN link (perhaps 1Mb/s), clients can quickly
saturate the WAN link. It is possible to configure the BITS client to never exceed a set threshold (e.g.
20Kbps) between two specific times in the day, but this does not allow BITS to take advantage of
available bandwidth beyond this threshold during quieter periods.



Is BranchCache right for remote, serverless software distribution?
Conclusion       Page 5



As noted previously, BranchCache will allocate 5% (default) of the disk capacity for use. When
the cache is first used, we observed 50% of the allocated space being „used‟, regardless of the
size of the content. If the BranchCache becomes full, older content will be removed to make
room for new content.

The BranchCache is separate from the ConfigMgr client cache. When using BranchCache with
ConfigMgr for software distribution it is therefore necessary to ensure there is sufficient disk
capacity for the content to be cached twice on each client (once in the BranchCache, once in the
ConfigMgr cache). In practice, observed behaviour 3 suggests the ConfigMgr cache is populated
from either the BranchCache on the local client or from the BranchCache on other clients, so the
total capacity required on each client would not always be double the size of the content, but
this must be the assumption when planning capacity.

Conclusion
For organizations that have Windows Server 2008 R2 and Windows 7 deployed throughout their
environment, BranchCache operating in Distributed Cache mode offers an ideal solution for ad-
hoc content retrieval from remote servers, such as obtaining documents from a file share or
SharePoint site, accessing web content or running a non-mandatory (i.e. unscheduled)
ConfigMgr Advertised Program without the need for a local server. In these scenarios it is very
unusual for several clients to simultaneously request the same content.

In the scenario where ConfigMgr is used to schedule the simultaneous deployment of an
application, patch or even OS image to a number of machines, our biggest concern was the
potential to saturate the network, as it was observed that each client started obtaining
significant amounts of content from the remote server before eventually settling down and
retrieving remaining content from peer caches. The optimum solution will only download content
once across the WAN, but our tests suggested that at least some of the content was downloaded
by several clients if the same content is requested simultaneously.

Our tests indicated that once content had been cached in a given subnet, as it was distributed
across a number of clients there was a higher chance that later requests for the same content
would result in some data being downloaded across the WAN from the remote server if some of
the original clients were powered off or removed – perhaps the reason that Microsoft does not
recommend using Distributed Mode in locations that include laptops.

Finally, if updates to the branch are frequent, or include OS images for rebuilding faulty
workstations, the amount of disk space required on each workstation when used with ConfigMgr
can be significant. Content is aged out of the BranchCache as capacity is required, so
applications that are only required occasionally in a branch may not be available even if they
were previously cached.




3
 We observed that with any particular package deployment, the volume of data in each participating
BranchCache was different, whereas the volume of data in each participating ConfigMgr cache was
consistent. This suggests that the ConfigMgr cache was populated directly from peer BranchCaches as
well as from the BranchCache on the local client.



Is BranchCache right for remote, serverless software distribution?

More Related Content

What's hot

Introduction to storage technologies
Introduction to storage technologiesIntroduction to storage technologies
Introduction to storage technologiesNuno Alves
 
Sunserver Open Solaris
Sunserver Open SolarisSunserver Open Solaris
Sunserver Open Solarispankaj009
 
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer Handbuch
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer HandbuchClearswift-Secure-Web-Gateway-Webfilter Load Balancer Handbuch
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer HandbuchLoadbalancer_org_Gmbh
 
Hyper v and failover clustering mini poster
Hyper v and failover clustering mini posterHyper v and failover clustering mini poster
Hyper v and failover clustering mini posterChawin Patimapornchai
 
Mysql wp memcached
Mysql wp memcachedMysql wp memcached
Mysql wp memcachedkbour23
 
Windows server 2012 R2 private cloud virtualization and storage
Windows server 2012 R2 private cloud virtualization and storageWindows server 2012 R2 private cloud virtualization and storage
Windows server 2012 R2 private cloud virtualization and storageSathishkumar A
 
Windows 2008 basics
Windows 2008 basicsWindows 2008 basics
Windows 2008 basicsanilinvns
 
Ws 2012 feature comparison hyper v
Ws 2012 feature comparison hyper vWs 2012 feature comparison hyper v
Ws 2012 feature comparison hyper vRamakrishnan M
 
Postgres plus cloud_database_getting_started_guide
Postgres plus cloud_database_getting_started_guidePostgres plus cloud_database_getting_started_guide
Postgres plus cloud_database_getting_started_guideice1oog
 
Server 2008 r2 ppt
Server 2008 r2 pptServer 2008 r2 ppt
Server 2008 r2 pptRaj Solanki
 
Windows clustering and quorum basics
Windows clustering and quorum basicsWindows clustering and quorum basics
Windows clustering and quorum basicsHarsh Chawla
 
SQL Server 2008 Consolidation
SQL Server 2008 ConsolidationSQL Server 2008 Consolidation
SQL Server 2008 Consolidationwebhostingguy
 
Innodb 1.1 for mysql 5.5 user's guide
Innodb 1.1 for mysql 5.5 user's guideInnodb 1.1 for mysql 5.5 user's guide
Innodb 1.1 for mysql 5.5 user's guidemysqlpub
 
Gateway Service For NetWare
Gateway Service For NetWareGateway Service For NetWare
Gateway Service For NetWareNderitu Muriithi
 

What's hot (18)

Introduction to storage technologies
Introduction to storage technologiesIntroduction to storage technologies
Introduction to storage technologies
 
Sunserver Open Solaris
Sunserver Open SolarisSunserver Open Solaris
Sunserver Open Solaris
 
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer Handbuch
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer HandbuchClearswift-Secure-Web-Gateway-Webfilter Load Balancer Handbuch
Clearswift-Secure-Web-Gateway-Webfilter Load Balancer Handbuch
 
Hyper v and failover clustering mini poster
Hyper v and failover clustering mini posterHyper v and failover clustering mini poster
Hyper v and failover clustering mini poster
 
Mysql wp memcached
Mysql wp memcachedMysql wp memcached
Mysql wp memcached
 
Windows server 2012 R2 private cloud virtualization and storage
Windows server 2012 R2 private cloud virtualization and storageWindows server 2012 R2 private cloud virtualization and storage
Windows server 2012 R2 private cloud virtualization and storage
 
What is mod_perl?
What is mod_perl?What is mod_perl?
What is mod_perl?
 
netezza-pdf
netezza-pdfnetezza-pdf
netezza-pdf
 
Windows 2008 basics
Windows 2008 basicsWindows 2008 basics
Windows 2008 basics
 
Ws 2012 feature comparison hyper v
Ws 2012 feature comparison hyper vWs 2012 feature comparison hyper v
Ws 2012 feature comparison hyper v
 
WINDOWS SERVER 2008
WINDOWS SERVER 2008WINDOWS SERVER 2008
WINDOWS SERVER 2008
 
Postgres plus cloud_database_getting_started_guide
Postgres plus cloud_database_getting_started_guidePostgres plus cloud_database_getting_started_guide
Postgres plus cloud_database_getting_started_guide
 
FailOver Clustring
FailOver ClustringFailOver Clustring
FailOver Clustring
 
Server 2008 r2 ppt
Server 2008 r2 pptServer 2008 r2 ppt
Server 2008 r2 ppt
 
Windows clustering and quorum basics
Windows clustering and quorum basicsWindows clustering and quorum basics
Windows clustering and quorum basics
 
SQL Server 2008 Consolidation
SQL Server 2008 ConsolidationSQL Server 2008 Consolidation
SQL Server 2008 Consolidation
 
Innodb 1.1 for mysql 5.5 user's guide
Innodb 1.1 for mysql 5.5 user's guideInnodb 1.1 for mysql 5.5 user's guide
Innodb 1.1 for mysql 5.5 user's guide
 
Gateway Service For NetWare
Gateway Service For NetWareGateway Service For NetWare
Gateway Service For NetWare
 

Similar to Nomad and WAN caching appliances 1.6

Planning Optimal Lotus Quickr services for Portal (J2EE) Deployments
Planning Optimal Lotus Quickr services for Portal (J2EE) DeploymentsPlanning Optimal Lotus Quickr services for Portal (J2EE) Deployments
Planning Optimal Lotus Quickr services for Portal (J2EE) DeploymentsStuart McIntyre
 
Optimizing windows 8 for virtual desktops - teched 2013 Jeff Stokes
Optimizing windows 8 for virtual desktops - teched 2013 Jeff StokesOptimizing windows 8 for virtual desktops - teched 2013 Jeff Stokes
Optimizing windows 8 for virtual desktops - teched 2013 Jeff StokesJeff Stokes
 
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]Krisman Tarigan
 
New Features For Your Software Defined Storage
New Features For Your Software Defined StorageNew Features For Your Software Defined Storage
New Features For Your Software Defined StorageDataCore Software
 
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docxaulasnilda
 
#VirtualDesignMaster 3 Challenge 1 – James Brown
#VirtualDesignMaster 3 Challenge 1 – James Brown#VirtualDesignMaster 3 Challenge 1 – James Brown
#VirtualDesignMaster 3 Challenge 1 – James Brownvdmchallenge
 
Peer Cache for Configuration Manager clients
Peer Cache for Configuration Manager clientsPeer Cache for Configuration Manager clients
Peer Cache for Configuration Manager clientsKarthick Jokirathinam
 
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...Cloudian
 
PowerPoint PresentationThis section will include an online pre.docx
PowerPoint PresentationThis section will include an online pre.docxPowerPoint PresentationThis section will include an online pre.docx
PowerPoint PresentationThis section will include an online pre.docxChantellPantoja184
 
Squid proxy-configuration-guide
Squid proxy-configuration-guideSquid proxy-configuration-guide
Squid proxy-configuration-guidejasembo
 
Windows Server Deployment Proposal For Worldwide Advertising, In.docx
Windows Server Deployment Proposal For Worldwide Advertising, In.docxWindows Server Deployment Proposal For Worldwide Advertising, In.docx
Windows Server Deployment Proposal For Worldwide Advertising, In.docxambersalomon88660
 
Introduction to Java Enterprise Edition
Introduction to Java Enterprise EditionIntroduction to Java Enterprise Edition
Introduction to Java Enterprise EditionAbdalla Mahmoud
 
Study notes for CompTIA Certified Advanced Security Practitioner (ver2)
Study notes for CompTIA Certified Advanced Security Practitioner  (ver2)Study notes for CompTIA Certified Advanced Security Practitioner  (ver2)
Study notes for CompTIA Certified Advanced Security Practitioner (ver2)David Sweigert
 
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceQuick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceCloudian
 
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceNetwork Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceCloudian
 
Virtual Design Master Challenge 1 - Akmal
Virtual Design Master Challenge 1  - AkmalVirtual Design Master Challenge 1  - Akmal
Virtual Design Master Challenge 1 - Akmaltovmug
 

Similar to Nomad and WAN caching appliances 1.6 (20)

Manual Sophos
Manual SophosManual Sophos
Manual Sophos
 
Windows Server 2003
Windows Server 2003Windows Server 2003
Windows Server 2003
 
Planning Optimal Lotus Quickr services for Portal (J2EE) Deployments
Planning Optimal Lotus Quickr services for Portal (J2EE) DeploymentsPlanning Optimal Lotus Quickr services for Portal (J2EE) Deployments
Planning Optimal Lotus Quickr services for Portal (J2EE) Deployments
 
Nt1330 Unit 1
Nt1330 Unit 1Nt1330 Unit 1
Nt1330 Unit 1
 
Optimizing windows 8 for virtual desktops - teched 2013 Jeff Stokes
Optimizing windows 8 for virtual desktops - teched 2013 Jeff StokesOptimizing windows 8 for virtual desktops - teched 2013 Jeff Stokes
Optimizing windows 8 for virtual desktops - teched 2013 Jeff Stokes
 
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]
18587936 squid-proxy-configuration-guide - [the-xp.blogspot.com]
 
New Features For Your Software Defined Storage
New Features For Your Software Defined StorageNew Features For Your Software Defined Storage
New Features For Your Software Defined Storage
 
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx
1Running head WINDOWS SERVER DEPLOYMENT PROPOSAL2WINDOWS SE.docx
 
#VirtualDesignMaster 3 Challenge 1 – James Brown
#VirtualDesignMaster 3 Challenge 1 – James Brown#VirtualDesignMaster 3 Challenge 1 – James Brown
#VirtualDesignMaster 3 Challenge 1 – James Brown
 
Peer Cache for Configuration Manager clients
Peer Cache for Configuration Manager clientsPeer Cache for Configuration Manager clients
Peer Cache for Configuration Manager clients
 
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...
How to become cloud backup provider with Cloudian HyperStore and CloudBerry L...
 
PowerPoint PresentationThis section will include an online pre.docx
PowerPoint PresentationThis section will include an online pre.docxPowerPoint PresentationThis section will include an online pre.docx
PowerPoint PresentationThis section will include an online pre.docx
 
Installation Guide
Installation GuideInstallation Guide
Installation Guide
 
Squid proxy-configuration-guide
Squid proxy-configuration-guideSquid proxy-configuration-guide
Squid proxy-configuration-guide
 
Windows Server Deployment Proposal For Worldwide Advertising, In.docx
Windows Server Deployment Proposal For Worldwide Advertising, In.docxWindows Server Deployment Proposal For Worldwide Advertising, In.docx
Windows Server Deployment Proposal For Worldwide Advertising, In.docx
 
Introduction to Java Enterprise Edition
Introduction to Java Enterprise EditionIntroduction to Java Enterprise Edition
Introduction to Java Enterprise Edition
 
Study notes for CompTIA Certified Advanced Security Practitioner (ver2)
Study notes for CompTIA Certified Advanced Security Practitioner  (ver2)Study notes for CompTIA Certified Advanced Security Practitioner  (ver2)
Study notes for CompTIA Certified Advanced Security Practitioner (ver2)
 
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceQuick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Quick-Start Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
 
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage ServiceNetwork Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
Network Setup Guide: Deploying Your Cloudian HyperStore Hybrid Storage Service
 
Virtual Design Master Challenge 1 - Akmal
Virtual Design Master Challenge 1  - AkmalVirtual Design Master Challenge 1  - Akmal
Virtual Design Master Challenge 1 - Akmal
 

More from 1E: Software Lifecycle Automation

Desktop Reclaim: Automate Removal of Microsoft & Adobe Software
Desktop Reclaim: Automate Removal of Microsoft & Adobe SoftwareDesktop Reclaim: Automate Removal of Microsoft & Adobe Software
Desktop Reclaim: Automate Removal of Microsoft & Adobe Software1E: Software Lifecycle Automation
 
Automated and dynamic maintenance keeps your computers healthy and performing...
Automated and dynamic maintenance keeps your computers healthy and performing...Automated and dynamic maintenance keeps your computers healthy and performing...
Automated and dynamic maintenance keeps your computers healthy and performing...1E: Software Lifecycle Automation
 
Drowsy server reduces energy consumption in your server farm
Drowsy server reduces energy consumption in your server farmDrowsy server reduces energy consumption in your server farm
Drowsy server reduces energy consumption in your server farm1E: Software Lifecycle Automation
 
Determining an accurate picture of energy usage with variable tariffs
Determining an accurate picture of energy usage with variable tariffsDetermining an accurate picture of energy usage with variable tariffs
Determining an accurate picture of energy usage with variable tariffs1E: Software Lifecycle Automation
 
Empowering energy efficiency and sustainability in it for federal agencies
Empowering energy efficiency and sustainability in it for federal agenciesEmpowering energy efficiency and sustainability in it for federal agencies
Empowering energy efficiency and sustainability in it for federal agencies1E: Software Lifecycle Automation
 
Reduce, reclaim and recycle your software to deliver dramatic cost reductions
Reduce, reclaim and recycle your software to deliver dramatic cost reductionsReduce, reclaim and recycle your software to deliver dramatic cost reductions
Reduce, reclaim and recycle your software to deliver dramatic cost reductions1E: Software Lifecycle Automation
 
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...1E: Software Lifecycle Automation
 
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...1E: Software Lifecycle Automation
 
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...1E: Software Lifecycle Automation
 
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...1E: Software Lifecycle Automation
 

More from 1E: Software Lifecycle Automation (20)

Desktop Reclaim: Automate Removal of Microsoft & Adobe Software
Desktop Reclaim: Automate Removal of Microsoft & Adobe SoftwareDesktop Reclaim: Automate Removal of Microsoft & Adobe Software
Desktop Reclaim: Automate Removal of Microsoft & Adobe Software
 
Automated and dynamic maintenance keeps your computers healthy and performing...
Automated and dynamic maintenance keeps your computers healthy and performing...Automated and dynamic maintenance keeps your computers healthy and performing...
Automated and dynamic maintenance keeps your computers healthy and performing...
 
How healthy is your computer estate?
How healthy is your computer estate?How healthy is your computer estate?
How healthy is your computer estate?
 
User empowerment and choice in downloading software
User empowerment and choice in downloading softwareUser empowerment and choice in downloading software
User empowerment and choice in downloading software
 
Create your own enterprise apps store
Create your own enterprise apps storeCreate your own enterprise apps store
Create your own enterprise apps store
 
Maximize the efficiency of your server farm
Maximize the efficiency of your server farmMaximize the efficiency of your server farm
Maximize the efficiency of your server farm
 
Drowsy server reduces energy consumption in your server farm
Drowsy server reduces energy consumption in your server farmDrowsy server reduces energy consumption in your server farm
Drowsy server reduces energy consumption in your server farm
 
Yes to virtualization projects but dont virtualize waste
Yes to virtualization projects but dont virtualize wasteYes to virtualization projects but dont virtualize waste
Yes to virtualization projects but dont virtualize waste
 
Determining an accurate picture of energy usage with variable tariffs
Determining an accurate picture of energy usage with variable tariffsDetermining an accurate picture of energy usage with variable tariffs
Determining an accurate picture of energy usage with variable tariffs
 
Agents vs Agentless
Agents vs AgentlessAgents vs Agentless
Agents vs Agentless
 
Empowering energy efficiency and sustainability in it for federal agencies
Empowering energy efficiency and sustainability in it for federal agenciesEmpowering energy efficiency and sustainability in it for federal agencies
Empowering energy efficiency and sustainability in it for federal agencies
 
PXE Lot or PXE Lite
PXE Lot or PXE LitePXE Lot or PXE Lite
PXE Lot or PXE Lite
 
Bios configuration management
Bios configuration managementBios configuration management
Bios configuration management
 
Reduce, reclaim and recycle your software to deliver dramatic cost reductions
Reduce, reclaim and recycle your software to deliver dramatic cost reductionsReduce, reclaim and recycle your software to deliver dramatic cost reductions
Reduce, reclaim and recycle your software to deliver dramatic cost reductions
 
Pragmatic software governance
Pragmatic software governancePragmatic software governance
Pragmatic software governance
 
Empower your users and save millions
Empower your users and save millionsEmpower your users and save millions
Empower your users and save millions
 
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...
NOMAD ENTERPRISE & WAN CACHING APPLIANCES NETWORK OPTIMIZATION IN A CONFIGURA...
 
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...
SWEATING THE ASSET HOW IMPROVING THE HEALTH AND PERFORMANCE OF YOUR PCS CAN E...
 
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...
A WASTE OF TIME AND MONEY? HOW IMPROVING THE HEALTH OF YOUR DESKTOPS AND LAPT...
 
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...
STOPPING THE YO-YO WHY CENTRALIZED DECENTRALIZATION IS THE KEY TO CUTTING COS...
 

Recently uploaded

08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024BookNet Canada
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationRadu Cotescu
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
Maximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxMaximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxOnBoard
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Alan Dix
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Igalia
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Servicegiselly40
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationSafe Software
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...HostedbyConfluent
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Drew Madelung
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 

Recently uploaded (20)

08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
Maximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptxMaximizing Board Effectiveness 2024 Webinar.pptx
Maximizing Board Effectiveness 2024 Webinar.pptx
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
 
CNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of ServiceCNv6 Instructor Chapter 6 Quality of Service
CNv6 Instructor Chapter 6 Quality of Service
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 

Nomad and WAN caching appliances 1.6

  • 1. Is BranchCache right for remote, serverless software distribution? 1E Technical Whitepaper Microsoft BranchCache and System Center Configuration Manager 2007 Abstract BranchCache is a new feature available in Windows Server 2008 R2 and Windows 7 that reduces WAN bandwidth usage and improves application responsiveness when workstations in a remote location access content from the head office or data center by downloading and caching content on the local network as it is requested, making it immediately available to other clients that subsequently request the same content. This paper examines the BrachCache functionality specifically in the context of software distribution using System Center Configuration Manager 2007 to determine if it is an optimal solution for the deployment of software, patches and operating systems to remote, serverless branches. Key Findings  BranchCache requires content to be hosted on Windows Server 2008 R2 and participating clients to have Windows 7 or Windows Vista with BITS 4.0 installed  BranchCache does not require a local server when operated in Distributed Cache mode, which is ideal for environments of around 50 clients. Microsoft recommend the Hosted Cache mode (which requires a local Windows Server 2008 R2 server in the branch) for remote locations that include laptops  BranchCache is a very efficient method for distributing „ad-hoc‟ content, including software distributions and updates that are initiated by the user at an arbitrary time (i.e. non-mandatory Advertisements that do not have a scheduled install time).  When the same content is requested simultaneously by several clients (as is the case with scheduled software and patch deployments), some of the content may be downloaded across the WAN by several clients. If the content is HTTP, BITS is used which may result in link saturation when multiple clients simultaneously download
  • 2. Is BranchCache right for Software Distribution? Copyright © 1E LTD 2011 Version 1.0 document revision 1 Author: Simon Burbidge & Dave Fuller All rights reserved. No part of this document or of the software (“the software”) to which it relates shall be reproduced, adapted, stored in a retrieval system, or transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise, without permission from 1E Ltd. It is the responsibility of the user to use the software in accordance with this document and 1E Ltd shall not be responsible if the user fails to do so. Although every precaution has been taken in the preparation of this document, 1E Ltd and the authors assume no responsibility for errors or omissions, nor shall they be liable for damages resulting from any information in it. Trademarks 1E is a trademark of 1E Ltd registered in the UK, with registration applied for in the US and EU. The 1E device is a trademark of 1E Ltd with registration applied for in the UK, US, EU and Australia. NIGHTWATCHMAN is a trademark of 1E Ltd registered in the US, with registration applied for in the EU and Australia. 1E WAKEUP is a trademark of 1E Ltd. MICROSOFT, WINDOWS SERVER 2008, WINDOWS VISTA, WINDOWS XP, WINDOWS 2000 and BRANCHCACHE are all trademarks of Microsoft Corporation in the United States and other countries.
  • 3. Contents Page i Contents Introduction 1 How does BranchCache work? 1 BranchCache Modes .......................................................................................................1 Discovery and Download .................................................................................................1 Caching 2 Setup and configuration ..................................................................................................2 The Remote Branch Software Distribution Challenge 2 How does BranchCache measure up? 3 No requirement for a local server .....................................................................................3 Efficient use of available bandwidth ..................................................................................3 No dependency for any one client to be available at all times ...............................................4 Efficient use of available disk capacity on local workstations .................................................4 Conclusion 5 Is BranchCache right for remote, serverless software distribution?
  • 4. Introduction Page 1 Introduction BranchCache is a new feature of Microsoft Windows Server 2008 R2 and Windows 7 (also available on Windows Vista with BITS 4.0 installed). It provides a software-based WAN optimization solution for branch environments enabling remote content requested by clients in the branch to be cached locally, either on a dedicated server or among peer workstations, so that it is available locally to other clients in the branch that may require the same content later rather than having them download it again over the WAN. Once set up, BranchCache is transparent to the application and is used to manage HTTP, HTTPS and SMB2 traffic, so any applications that use these protocols can benefit from BranchCache. Microsoft included BranchCache support in the SP2 release of System Center Configuration Manager (ConfigMgr) 2007, enabling its software distribution, software updates and image deployment features to take advantage of this feature in environments where Windows 2008 R2 and Windows 7 have been deployed and the BranchCache feature configured. This paper takes a look at the BranchCache feature in detail, focusing on how the feature performs in remote software distribution, patch management and image deployment scenarios. Although BranchCache can be configured to use a local Windows Server 2008 R2 server to store the cache (explained in BranchCache Modes), this paper focuses on the server-less Distributed Mode, on the assumption that if a local server were available, it would make sense to host a ConfigMgr Secondary Site or Distribution Point on that server anyway. How does BranchCache work? BranchCache Modes BranchCache can operate in one of two modes: Hosted Mode or Distributed Mode. Hosted Mode requires a local Windows Server 2008 R2 server in the branch and clients to be configured with the FQDN of this server. When a client requests content from a remote server that is not already present on the local BranchCache server, the client downloads the content but then it is moved to the local server (the content is actually advertised to the server and the server downloads it from the client). The local cache is maintained on the server and subsequent requests for the same content from other clients in the branch can be served from this cache. Distributed Mode does not require a local server. Instead, content is cached on the workstation clients and is shared with peer clients when they subsequently require the same content (remember only Windows 7 or Windows Vista with BITS 4.0 clients can participate). Distributed Mode is suitable for smaller environments (Microsoft recommends up to 50 clients). Clients participating in peer-to-peer content transfer must be in the same subnet. Microsoft recommends using Hosted Mode for branches with more than 50 client and branches that have laptops that may be removed from the network. Discovery and Download BranchCache is transparent to the application. When the application running on a BranchCache enabled client requests content from a BranchCache enabled server over HTTP/HTTPS or SMB2, the BranchCache client will request content metadata from the remote server. Metadata is generated on the source server each time it is requested by the client and consists of block hashes, segment hashes and a segment key, a segment being a collection of blocks. Typically the hashing results in a compression ratio of 2000:1, meaning that the metadata is typically 1/2000 (or 0.05%) of the size of the source data. A BranchCache client will always request this metadata from the source server regardless of whether the data already exists locally to ensure that the source content has not changed. Once the metadata has been downloaded, in Distributed Mode clients perform a local discovery (using the Branch Cache Discovery Protocol, which in turn uses the WS-Discovery protocol) to determine if any of the segments are available within the branch before resorting to downloading these from the remote server. HTTP/HTTPS content is downloaded using Background Intelligent Transfer Service (BITS). Interestingly, when using SMB2, BranchCache can be configured to be invoked according to connection latency (a threshold set using NETSH) so allowing BranchCache to only be used Is BranchCache right for remote, serverless software distribution?
  • 5. The Remote Branch Software Distribution Challenge Page 2 when the connection to the server is performing poorly. In contrast, BranchCache using BITS does not possess this intelligence and is always invoked if enabled. Note that BranchCache does not download files smaller than 64KB (the requesting application will receive these through normal SMB2 / HTTP transfer). Caching When BranchCache is implemented in Distributed Mode, each client is allocated a proportion of the total (not free) disk size. By default this is 5% but can be adjusted using NETSH. When the BranchCache service first starts, the cache is initialized with sparse files (0 data files with a .PDS extension). These will eventually host the downloaded package and hash data. During testing it was seen that when BranchCache is actually used for the first time, it appears to instantly reserve half its cache allocation with the 0 byte data .PDS file being expanded to 50% allocated cash size (regardless of the size of the actual package being downloaded). It should be noted that the BranchCache is completely independent from the ConfigMgr client cache. The ConfigMgr client will populate its own cache from wherever BranchCache gets the content from. Using BranchCache with ConfigMgr therefore requires sufficient disk space on the client to host the content in both the BranchCache and the ConfigMgr cache. Setup and configuration As long as your source servers are running Windows Server 2008 R2 and your workstations are on Windows 7 (or Vista with BITS 4.0), setting BranchCache up is relatively straightforward. BranchCache is installed as a Feature in the Server Manager interface. If you want to support file transfer (SMB2) with BranchCache you need to add the BranchCache for network files role in the File Services server role (also managed through Server Manager). On clients, BranchCache is disabled by default. Most organisations would use Group Policy to enable and configure BranchCache, typically defining the cache mode, size and location. BranchCache can also be enabled and managed using the NETSH command-line utility. The NETSH command-line tool can also be used to manage the local cache on a client (such as flushing it) and can be used to display configuration settings and cache content. The Remote Branch Software Distribution Challenge Remote branch offices will typically range from a handful of clients to perhaps a couple of hundred. Often these will be a healthy mix of desktops and laptops, with some laptops being off the network for periods of the time. In many cases there is no facility (or desire) to host servers in these locations, and often network bandwidth to the head offices and datacenters is highly contested. The distribution of software (applications, patches and OS images) in these locations is somewhat different to the user-initiated applications (e.g. accessing documents from a SharePoint server) commonly used to demonstrate BranchCache functionality. Many software and patch deployments require content to be distributed to all machines in the branch at the same time (typically the case with Software Updates, but also with the planned roll-out of a new application). Software applications and updates may be distributed in the form of a single „self-extracting‟ file (e.g. in MSI or EXE format), or may involve the distribution of several files. Distribution of OS images requires the transfer of very large WIM files and the process of deploying the image to a workstation will typically require additional package content (applications) not included in the image. These deployments have to contend with other applications using the WAN and must avoid adversely impacting the normal flow of business relying upon it. In order to achieve deployment within an acceptable time period, WAN traffic should be minimised and bandwidth used efficiently. This involves ensuring the content only traverses the WAN once and is shared out within the branch (rather than each client obtaining it from the source server), and ensuring the download process is aware of existing traffic levels and is able to throttle demand on the link accordingly, taking advantage of „quiet‟ periods but backing off when other traffic requires the bandwidth. In a serverless branch, the dependency is on the client workstations to provide the local cache. It is therefore important that the workstation that hosts the cache is available at all times, or that the cache is replicated on several machines to provide resilience. Is BranchCache right for remote, serverless software distribution?
  • 6. How does BranchCache measure up? Page 3 The ideal solution for software distribution to remote branches would have the following attributes  No requirement for a local server  Efficient use of available bandwidth, making use of bandwidth as it becomes available but ensuring it is never saturated and ensuring content is only transferred to the branch once  No dependency for any one client to be available at all times  Efficient use of available disk capacity on local workstations How does BranchCache measure up? 1E conducted research into the use of BranchCache in the software distribution scenario, specifically with the use of System Center Configuration Manager (ConfigMgr) 2007 Service Pack 2, which introduced specific integration with BranchCache. This section summarises the results of our testing in relation to the ideal attributes listed above. No requirement for a local server Implementing BranchCache in Distributed mode supports this requirement. Windows 7 and Windows Vista clients with BITS v4.0 can participate in the peer-caching. Microsoft recommends that Distributed mode be used for branches up to 50 clients. If clients span multiple subnets then Distributed mode results in one download per subnet. In the Key BranchCache Design Points section of the Microsoft Windows Server 20087 R2 BranchCache Design Guide 1 it states that the BranchCache data can only be decrypted by authenticated clients that are members of the same domain as the content server, although we have not verified this requirement. Efficient use of available bandwidth When multiple clients request the same content simultaneously, as is the case when a deployment is scheduled at a specific time, we observed that all clients begin to download data from the source server simultaneously. (The BranchCache performance counters were used to measure the volume of data downloaded from the WAN, downloaded from a peer cache and served to others on the network). This is expected as every client must obtain the metadata from the source to ensure data integrity, and as previously noted this should amount to 0.05% of the content size if 2000:1 compression is achieved. However, we observed that overall, the amount of data downloaded across the WAN to the branch was between 2 and 11% greater than the content size (i.e. the overhead was between 2 and 11%), suggesting that clients were simultaneously downloading content as well as metadata. The tables below show some examples of our test results.  Bytes from server is the amount of data transferred across the WAN from the remote server  Bytes from Cache is the amount of data that the client obtained from the cache on another local workstation  Bytes Served is the amount of data that each client provided to other clients from its own cache 5MB Single file package deployment Client Bytes from Bytes from Bytes Served Server Cache A 204KB 4.8MB 1.6MB B 347KB 4.6MB 196KB C 5MB 0 7.6MB In this test, all clients began downloading from the server simultaneously. After about 20 seconds, A and B switch to peer caching. The bytes served figures indicates that A and B cached 1 http://www.microsoft.com/downloads/details.aspx?FamilyID=CBF75A21-BC09-4824-B128- 1A24EE71A9AA&displayLang=en Is BranchCache right for remote, serverless software distribution?
  • 7. How does BranchCache measure up? Page 4 from C as well as from each other. As noted earlier, the file metadata (data that must be first downloaded from the server) should be using a compression ratio of 2000:1 (0.05% of file size) or 2.5KB for the test example above. The results show that clients downloaded substantially more than this amount from the server, suggesting that either the desired compression ratio was not achieved, or clients were downloading content as well as metadata from the server, which is inefficient. 15MB package containing 10 unique files Client Bytes from Bytes from Bytes Served Server Cache A 5.1MB 7.9MB 9.9MB B 8MB 6MB 13.5MB C 2.2MB 12MB 2.5MB In this scenario it is obvious that the clients are simultaneously downloading significant amounts of content directly from the server. As these clients are using BITS to download the content, it is not possible to dynamically make optimum use of the available WAN bandwidth 2, and as more clients participate in the download, the chances of saturating the WAN link are increased. 5MB 10 file package (all files had identical content) Client Bytes from Bytes from Bytes Served Server Cache A 4KB 5MB 0 B 500KB 3.8MB 458KB C 570KB 3.2MB 500KB The figures illustrate an interesting feature of BranchCache. Although we have a 5MB package, we have only downloaded a total of approximately 1MB from the server. The Program ran successfully with the entire package present within the ConfigMgr client cache. BranchCache was able to recognise the fact that the all files shared the same content and effectively reused data when supplying the clients. BranchCache data discovery behaviour further illustrates this point. Within this scenario, despite there being 10 files in the source, clients issued only 2 discovery requests as opposed when dealing with unique files, when a discovery request appears to be issued per file. No dependency for any one client to be available at all times The BranchCache Distributed cache model, it was observed that not all participating clients do not cache the entire content. During a test deployment of a 2.7GB OS image to 3 computers, only one client had the full 2.7GB in the BranchCache (although the ConfigMgr client cache had the whole content) while the others had around 1.3GB. The client that had the full image in its BranchCache was then switched off and the image deployed to another client in the subnet. This client had to obtain the balance of the image (1.4GB) from the server. The other tests described in the previous section also revealed that clients had different volumes of data in their cache, indicating that the caching is indeed distributed and that in order for other clients to obtain all the content during a subsequent deployment, multiple clients would need to be available on the network to enable the client to obtain the total content from the local network. Efficient use of available disk capacity on local workstations Although hard disk capacity is pretty cheap these days, using workstation-class hardware to store distributed content should still make efficient use of the available space, especially if the workstation is being used to perform a user‟s day job as well. 2 The BITS client makes use of bandwidth available on the local network adapter and is not aware of the end-to-end bandwidth available on the WAN. As the local network adapter typically has significantly more bandwidth (e.g. 100Mb/s) than the WAN link (perhaps 1Mb/s), clients can quickly saturate the WAN link. It is possible to configure the BITS client to never exceed a set threshold (e.g. 20Kbps) between two specific times in the day, but this does not allow BITS to take advantage of available bandwidth beyond this threshold during quieter periods. Is BranchCache right for remote, serverless software distribution?
  • 8. Conclusion Page 5 As noted previously, BranchCache will allocate 5% (default) of the disk capacity for use. When the cache is first used, we observed 50% of the allocated space being „used‟, regardless of the size of the content. If the BranchCache becomes full, older content will be removed to make room for new content. The BranchCache is separate from the ConfigMgr client cache. When using BranchCache with ConfigMgr for software distribution it is therefore necessary to ensure there is sufficient disk capacity for the content to be cached twice on each client (once in the BranchCache, once in the ConfigMgr cache). In practice, observed behaviour 3 suggests the ConfigMgr cache is populated from either the BranchCache on the local client or from the BranchCache on other clients, so the total capacity required on each client would not always be double the size of the content, but this must be the assumption when planning capacity. Conclusion For organizations that have Windows Server 2008 R2 and Windows 7 deployed throughout their environment, BranchCache operating in Distributed Cache mode offers an ideal solution for ad- hoc content retrieval from remote servers, such as obtaining documents from a file share or SharePoint site, accessing web content or running a non-mandatory (i.e. unscheduled) ConfigMgr Advertised Program without the need for a local server. In these scenarios it is very unusual for several clients to simultaneously request the same content. In the scenario where ConfigMgr is used to schedule the simultaneous deployment of an application, patch or even OS image to a number of machines, our biggest concern was the potential to saturate the network, as it was observed that each client started obtaining significant amounts of content from the remote server before eventually settling down and retrieving remaining content from peer caches. The optimum solution will only download content once across the WAN, but our tests suggested that at least some of the content was downloaded by several clients if the same content is requested simultaneously. Our tests indicated that once content had been cached in a given subnet, as it was distributed across a number of clients there was a higher chance that later requests for the same content would result in some data being downloaded across the WAN from the remote server if some of the original clients were powered off or removed – perhaps the reason that Microsoft does not recommend using Distributed Mode in locations that include laptops. Finally, if updates to the branch are frequent, or include OS images for rebuilding faulty workstations, the amount of disk space required on each workstation when used with ConfigMgr can be significant. Content is aged out of the BranchCache as capacity is required, so applications that are only required occasionally in a branch may not be available even if they were previously cached. 3 We observed that with any particular package deployment, the volume of data in each participating BranchCache was different, whereas the volume of data in each participating ConfigMgr cache was consistent. This suggests that the ConfigMgr cache was populated directly from peer BranchCaches as well as from the BranchCache on the local client. Is BranchCache right for remote, serverless software distribution?