Apache Wiki | 2 Sep 2011 14:20
Picon
Favicon

[Httpd Wiki] Trivial Update of "ScratchPad" by wrowe

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "ScratchPad" page has been changed by wrowe:
http://wiki.apache.org/httpd/ScratchPad?action=diff&rev1=40&rev2=41

Comment:
Drop conflict tags

  
  Multi-domain, SAN, or UCC certificates are useful when organizations require different root domain
names to run Internet-facing services.  Subject alternate name certificates are also called Unified
Communications Certificates (UCC) since they were primarily designed to support real-time
communications infrastructures. For example, an organization providing both internal and external
unified communications services with two different domain names—for example,
SIP.VirtualSpaceShip.com the external domain and SIP.VirtualSpaceShip.com for the internal
name—would benefit from a multi-domain certificate because in this case, the wildcard certificate
would not work. In fact, if the organization was using wildcard certificates, two wildcard certificates
would be required because the root domain name is different in each case.

- 
- ---- /!\ '''Edit conflict - other version:''' ----
  A single multi-domain certificate could easily support the following names and more:
www.VirtualSpaceShip.com, www.VirtualSpaceShip.ws, www.VirtualCDVD.com,
www.VirtualWorkersInSpace.com 
- 
- ---- /!\ '''Edit conflict - your version:''' ----
- A single multi-domain certificate could easily support the following names and more:
www.VirtualSpaceShip.com, www.VirtualSpaceShip.ws, www.VirtualCDVD.com,
(Continue reading)

Geoffrey Noakes | 2 Sep 2011 18:53
Favicon

VeriSign documentation choosing between SAN/Wildcard certs in the Apache HTTPD environment (resend)

HTTPD group, Don and I are with Symantec’s Trust Services group – which is probably better known to the Apache community as VeriSign SSL (which includes GeoTrust and Thawte).  We have been working with Bill Rowe on this. 

 

We have uploaded 3+ screens of content about choosing between SAN and Wildcard certs to http://wiki.apache.org/httpd/ScratchPad.  We have more content, primarily in the form of tables, and we didn't know how to include that in the Wiki.  This is *not* meant to be a product recommendation piece, instead it is an education paper.

 

What we would like to do is to work with Apache’s HTTP Community to complete this work.  We are specifically asking for anyone who is particularly experienced with SSL to comment on it or offer edits.

 

Please advise us on how best to move forward and start a dialog about incorporating this scratchpad draft into the Apache documentation.

 

Thanks...

 

Geoff

 

Geoffrey W. Noakes

Director, Business Development

Symantec Corporation

geoffrey_noakes <at> symantec.com

+1-415-370-5980

 

 

 

William A. Rowe Jr. | 2 Sep 2011 19:08

Re: VeriSign documentation choosing between SAN/Wildcard certs in the Apache HTTPD environment (resend)

On 9/2/2011 11:53 AM, Geoffrey Noakes wrote:
> 
> We have uploaded 3+ screens of content about choosing between SAN and Wildcard certs to
> http://wiki.apache.org/httpd/ScratchPad.  We have more content, primarily in the form of
> tables, and we didn't know how to include that in the Wiki.

Geoff, the tabular data can be formatted with

|| col1 || col2 ||

see http://wiki.apache.org/httpd/HelpOnMoinWikiSyntax#Tables
Apache Wiki | 9 Sep 2011 12:14
Picon
Favicon

[Httpd Wiki] Trivial Update of "ScratchPad" by jmcg

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "ScratchPad" page has been changed by jmcg:
http://wiki.apache.org/httpd/ScratchPad?action=diff&rev1=41&rev2=42

Comment:
trival replacements of weird characters.

  Understanding Multi-Use Digital Certificates

  Abstract
- With the proliferation of internal Web-based services that must be exposed to the Internet,
organizations are turning more and more to the use of SSL certificates. However, using traditional SSL
certificates can become cumbersome and quite expensive, especially when organizations offer several
Internet-facing services. This is where multi-use certificates—certificates that can be reused for
different purposes—can help. Using a single multi-use certificate, organizations can reduce costs
and simplify certificate management.
+ With the proliferation of internal Web-based services that must be exposed to the Internet,
organizations are turning more and more to the use of SSL certificates. However, using traditional SSL
certificates can become cumbersome and quite expensive, especially when organizations offer several
Internet-facing services. This is where multi-use certificates -- certificates that can be reused for
different purposes -- can help. Using a single multi-use certificate, organizations can reduce costs
and simplify certificate management.

  "'Introduction'"
- The best way to prove who you are on the Internet is to use a digital certificate. That’s because a digital
certificate relies on a trusted, third-party authority to verify your identity. In fact, it uses a chain
of trust that begins with you and works its way up to the trusted authority that validates who you are. This
chain of trust provides verifiable Internet security.
+ The best way to prove who you are on the Internet is to use a digital certificate. That's because a digital
certificate relies on a trusted, third-party authority to verify your identity. In fact, it uses a chain
of trust that begins with you and works its way up to the trusted authority that validates who you are. This
chain of trust provides verifiable Internet security.
  Take Internet user Bill for example. Bill uses a digital certificate to sign all of his emails. This example
demonstrates the concept of authenticity. The certificate authenticates Bill as the author of the email.
- The digital certificate also verifies that Bill is the actual author and sender of the email. This is the
concept of non-repudiation—by using a certificate, you can certify with reasonable certainty that the
signed document is trusted to be from Bill or at least someone who possesses the private key corresponding
to the signing certificate.
+ The digital certificate also verifies that Bill is the actual author and sender of the email. This is the
concept of non-repudiation -- by using a certificate, you can certify with reasonable certainty that the
signed document is trusted to be from Bill or at least someone who possesses the private key corresponding
to the signing certificate.
  In addition, because Bill signs his emails with a digital certificate, the email cannot be tampered with
without invalidating the signature. This is the concept of integrity. Because the email includes a
digital signature, it cannot be modified while in transit without the tampering being obvious to the recipient.
- Two more security concepts are supported by digital certificates. Availability is critical to the
third-party organization that certifies that Bill is who he claims to be. The verification service must
be available when you need to verify Bill’s certificate. Finally, confidentiality is provided by the
ability to encrypt data both in the content itself and in the transport mechanism that is used to send the
data to a destination on the Internet.
+ Two more security concepts are supported by digital certificates. Availability is critical to the
third-party organization that certifies that Bill is who he claims to be. The verification service must
be available when you need to verify Bill's certificate. Finally, confidentiality is provided by the
ability to encrypt data both in the content itself and in the transport mechanism that is used to send the
data to a destination on the Internet.

- Similarly, organizations that process transactions on the Internet or that offer Internet-based
services need to rely on digital certificates to validate that they are who they claim to be; otherwise, no
one will trust their services. Most organizations do this by adding certificates to their
Internet-facing servers. When users access a web page hosted on one of these servers, their Web browser
will automatically detect the certificate and modify the session, from an open session using the
HyperText Transfer Protocol (HTTP) to secure HTTP (HTTPS).   This will allow for the encryption of all the
data sent between the user’s workstation and the server. HTTPS data encryption is provided by the Secure
Sockets Layer (SSL). Basically, SSL creates an encryption tunnel between the client and the server
protecting the transfer of data from one point to the other during the communication exchange. You know
you are using SSL when your browser displays a closed padlock in its status bar. 
+ Similarly, organizations that process transactions on the Internet or that offer Internet-based
services need to rely on digital certificates to validate that they are who they claim to be; otherwise, no
one will trust their services. Most organizations do this by adding certificates to their
Internet-facing servers. When users access a web page hosted on one of these servers, their Web browser
will automatically detect the certificate and modify the session, from an open session using the
HyperText Transfer Protocol (HTTP) to secure HTTP (HTTPS).   This will allow for the encryption of all the
data sent between the user's workstation and the server. HTTPS data encryption is provided by the Secure
Sockets Layer (SSL). Basically, SSL creates an encryption tunnel between the client and the server
protecting the transfer of data from one point to the other during the communication exchange. You know
you are using SSL when your browser displays a closed padlock in its status bar. 

  "'Working with SSL Certificates'"
- In order to enable SSL on their external-facing servers, organizations must purchase a certificate from
a trusted certification authority for each protected service they provide. Today, organizations can
offer several services to their end users—email, instant messaging, mobile device management,
Web-based interactions and more—and each one of the servers providing these services requires its own
certificate. 
+ In order to enable SSL on their external-facing servers, organizations must purchase a certificate from
a trusted certification authority for each protected service they provide. Today, organizations can
offer several services to their end users -- email, instant messaging, mobile device management,
Web-based interactions and more -- and each one of the servers providing these services requires its own
certificate. 

  SSL certificates are tied to the unique domain name on which a service is hosted. Embedding the domain name
into the certificate is important since this makes it possible to ensure the identity of the remote
computer providing the service by comparing the domain name being accessed to the domain name included in
the certificate itself.
  Certificates are issued for a finite period, usually in 12-month increments. Because of this, you should
aim to obtain certificates that are valid for extended periods of time. You should also aim to select
static service names because each time a service name changes, the certificate must be changed on each
server that provides the service. These strategies will reduce the workload associated with periodic
renewal and installation of certificates on your servers. 
- These problems are compounded when organizations must use different domain names for each secure
service they make available on the Internet. In fact, organizations often find themselves in a situation
where they need to use sub-domain names—names that use the same root name, but require a different prefix
name—to secure each of the services they offer. Because prefix names are embedded into SSL
certificates, organizations usually buy one certificate per service. As you can imagine, this can
become expensive and time-consuming to manage, especially in organizations that run a multitude of
Internet-facing services. 
+ These problems are compounded when organizations must use different domain names for each secure
service they make available on the Internet. In fact, organizations often find themselves in a situation
where they need to use sub-domain names -- names that use the same root name, but require a different prefix
name -- to secure each of the services they offer. Because prefix names are embedded into SSL
certificates, organizations usually buy one certificate per service. As you can imagine, this can
become expensive and time-consuming to manage, especially in organizations that run a multitude of
Internet-facing services. 
  Enter the multi-use SSL certificate. There are two types of multi-use certificates:
  Wildcard certificates can secure multiple sub-domains on a single unique Fully Qualified Domain Name
using a single certificate.
  Multi-domain certificates can secure multiple Fully Qualified Domain Names using a single certificate. 
  Each certificate can simplify management and reduce costs given the right situation.

  "'Wildcard Certificates'"
- The first type of multi-use certificate is the wildcard certificate. The name you embed in a certificate
must always follow the fully-qualified domain name (FQDN) format. If you want a certificate for the
Session Initiation Protocol (SIP) used by instant communication servers in the VirtualSpaceShip.com
domain, the name embedded into the certificate will be SIP.VirtualSpaceShip.com. If you want a
certificate for the email service, then you would normally have to buy a second certificate with the
second service name—mail.VirtualSpaceShip.com—embedded into it.
+ The first type of multi-use certificate is the wildcard certificate. The name you embed in a certificate
must always follow the fully-qualified domain name (FQDN) format. If you want a certificate for the
Session Initiation Protocol (SIP) used by instant communication servers in the VirtualSpaceShip.com
domain, the name embedded into the certificate will be SIP.VirtualSpaceShip.com. If you want a
certificate for the email service, then you would normally have to buy a second certificate with the
second service name -- mail.VirtualSpaceShip.com -- embedded into it.
  In addition, some secure service implementations require internal as well as external validation and you
use a different name for each; for example, InternalSip.VirtualSpaceShip.com  and
ExternalSIP.VirtualSpaceShip.com. In this case, you must have a certificate on each server in the
internal and external service to allow users to work unimpeded whether they are in the office or on the
road. This is the case for instant messaging infrastructures where you want to ensure messages are
encrypted whether they are internal or external. Note that servers cannot include two certificates for
the same purpose.
  Wildcard certificates do not include service names. Instead, they are standard certificates that
support the use of a wildcard character to replace the prefix name in the subject name field, for example,
*.VirtualSpaceShip.com.  Using a wildcard certificate is much more practical and versatile than using
multiple single purpose certificate since the wildcard certificate can be applied to a number of
different services without requiring any updates. In addition, you can add, change or replace services
without needing to update the certificate. 

 <at>  <at>  -39, +39  <at>  <at> 

  Multi-domain certificates include the standard Subject Name field which supports a single primary
service name, as well as an additional entry called the Subject Alternative Name field which supports the
additional service names. The SAN certificate can therefore be installed on several servers and
function properly to support internal/external service delivery.
  SAN certificates have the same issues as single-purpose certificates however. Because the actual
service names are embedded into the certificate, you must make sure your services always use the same name
otherwise you must change the certificate and since the certificate is a multi-use certificate, you must
change it on each of the computers that host the service which the certificate supports. Additionally,
when you want to add services to provide further functionality to your users, you must update the SAN
certificate with the new service names.

- Multi-domain, SAN, or UCC certificates are useful when organizations require different root domain
names to run Internet-facing services.  Subject alternate name certificates are also called Unified
Communications Certificates (UCC) since they were primarily designed to support real-time
communications infrastructures. For example, an organization providing both internal and external
unified communications services with two different domain names—for example,
SIP.VirtualSpaceShip.com the external domain and SIP.VirtualSpaceShip.com for the internal
name—would benefit from a multi-domain certificate because in this case, the wildcard certificate
would not work. In fact, if the organization was using wildcard certificates, two wildcard certificates
would be required because the root domain name is different in each case.
+ Multi-domain, SAN, or UCC certificates are useful when organizations require different root domain
names to run Internet-facing services.  Subject alternate name certificates are also called Unified
Communications Certificates (UCC) since they were primarily designed to support real-time
communications infrastructures. For example, an organization providing both internal and external
unified communications services with two different domain names -- for example,
SIP.VirtualSpaceShip.com the external domain and SIP.VirtualSpaceShip.com for the internal name --
would benefit from a multi-domain certificate because in this case, the wildcard certificate would not
work. In fact, if the organization was using wildcard certificates, two wildcard certificates would be
required because the root domain name is different in each case.

  A single multi-domain certificate could easily support the following names and more:
www.VirtualSpaceShip.com, www.VirtualSpaceShip.ws, www.VirtualCDVD.com,
www.VirtualWorkersInSpace.com 

- Multi-domain certificates are also useful for application service providers (ASP) who host
applications for multiple clients with each client using their own domain name. By using a multi-domain
certificate, ASPs can use a single certificate to support multiple clients. Note that the site seal and
certificate “Issued To” will only be for the primary domain name entered in the certificate and will not
include any of the other domain names. However, the certificate itself will include all of the domain
names that have been entered when the certificate was purchased. 
+ Multi-domain certificates are also useful for application service providers (ASP) who host
applications for multiple clients with each client using their own domain name. By using a multi-domain
certificate, ASPs can use a single certificate to support multiple clients. Note that the site seal and
certificate "Issued To" will only be for the primary domain name entered in the certificate and will not
include any of the other domain names. However, the certificate itself will include all of the domain
names that have been entered when the certificate was purchased. 

  While multi-domain certificates are also useful when used to support unified communications
deployments, there are some caveats for their use:
  - Multi-domain certificates do not support use of wildcard characters.  For this reason, sub-domain names
must be added as a unique domain name entries in the certificate. Each time a new domain name is added or an
old one is removed the certificate must be updated and re-deployed to each host server.
 <at>  <at>  -56, +56  <at>  <at> 

  Both certificate types offer several benefits and include several features. In addition, both are
available in full authentication format only. Domain-only authentication certificates only require
the validation of the domain before they are issued, and because of this, cannot be used with multi-use
certificates. Full authentication certificates require both the validation of the domain itself and
the validation of the business running the domain. Because of this, full certificates are more
trustworthy than domain-only certificates. This is another reason the full authentication model is
used for multi-use certificates. Keep this in mind as you make your choice.

  "'Making the Selection'"
- Multi-use certificates were developed to provide multiple secure services originating from a single IP
address. To accomplish this, these certificates either add a subject alternate name (SAN) field to the
common single-use certificate or use a wildcard to replace the service name in the certificate.
Microsoft Exchange Server 2007 is an excellent example of this type of requirement. The same external
Exchange server can publish several different types of services: Outlook Web Access, Outlook Anywhere
Access, AutoDiscovery configuration information, and more. Each of these services requires the
publishing of its own name—for example, OWA.VirtualSpaceShip.com, Mail.VirtualSpaceShip.com,
Autodiscover.VirtualSpaceShip.com— ideally within a single certificate.
+ Multi-use certificates were developed to provide multiple secure services originating from a single IP
address. To accomplish this, these certificates either add a subject alternate name (SAN) field to the
common single-use certificate or use a wildcard to replace the service name in the certificate.
Microsoft Exchange Server 2007 is an excellent example of this type of requirement. The same external
Exchange server can publish several different types of services: Outlook Web Access, Outlook Anywhere
Access, AutoDiscovery configuration information, and more. Each of these services requires the
publishing of its own name -- for example, OWA.VirtualSpaceShip.com, Mail.VirtualSpaceShip.com,
Autodiscover.VirtualSpaceShip.com --  ideally within a single certificate.

  Multi-use certificates reduce cost and simplify management by supporting the inclusion of multiple
names within the same certificate or the replacement of service names with a wildcard. However, each
multi-use certificate type should only be used in specific situations:

Apache Wiki | 9 Sep 2011 12:19
Picon
Favicon

[Httpd Wiki] Update of "ScratchPad" by jmcg

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "ScratchPad" page has been changed by jmcg:
http://wiki.apache.org/httpd/ScratchPad?action=diff&rev1=42&rev2=43

  
  "'Introduction'"
  The best way to prove who you are on the Internet is to use a digital certificate. That's because a digital
certificate relies on a trusted, third-party authority to verify your identity. In fact, it uses a chain
of trust that begins with you and works its way up to the trusted authority that validates who you are. This
chain of trust provides verifiable Internet security.
+ 
  Take Internet user Bill for example. Bill uses a digital certificate to sign all of his emails. This example
demonstrates the concept of authenticity. The certificate authenticates Bill as the author of the email.
  The digital certificate also verifies that Bill is the actual author and sender of the email. This is the
concept of non-repudiation -- by using a certificate, you can certify with reasonable certainty that the
signed document is trusted to be from Bill or at least someone who possesses the private key corresponding
to the signing certificate.
+ 
  In addition, because Bill signs his emails with a digital certificate, the email cannot be tampered with
without invalidating the signature. This is the concept of integrity. Because the email includes a
digital signature, it cannot be modified while in transit without the tampering being obvious to the recipient.
+ 
  Two more security concepts are supported by digital certificates. Availability is critical to the
third-party organization that certifies that Bill is who he claims to be. The verification service must
be available when you need to verify Bill's certificate. Finally, confidentiality is provided by the
ability to encrypt data both in the content itself and in the transport mechanism that is used to send the
data to a destination on the Internet.

  Similarly, organizations that process transactions on the Internet or that offer Internet-based
services need to rely on digital certificates to validate that they are who they claim to be; otherwise, no
one will trust their services. Most organizations do this by adding certificates to their
Internet-facing servers. When users access a web page hosted on one of these servers, their Web browser
will automatically detect the certificate and modify the session, from an open session using the
HyperText Transfer Protocol (HTTP) to secure HTTP (HTTPS).   This will allow for the encryption of all the
data sent between the user's workstation and the server. HTTPS data encryption is provided by the Secure
Sockets Layer (SSL). Basically, SSL creates an encryption tunnel between the client and the server
protecting the transfer of data from one point to the other during the communication exchange. You know
you are using SSL when your browser displays a closed padlock in its status bar. 
 <at>  <at>  -19, +22  <at>  <at> 

  
  SSL certificates are tied to the unique domain name on which a service is hosted. Embedding the domain name
into the certificate is important since this makes it possible to ensure the identity of the remote
computer providing the service by comparing the domain name being accessed to the domain name included in
the certificate itself.
  Certificates are issued for a finite period, usually in 12-month increments. Because of this, you should
aim to obtain certificates that are valid for extended periods of time. You should also aim to select
static service names because each time a service name changes, the certificate must be changed on each
server that provides the service. These strategies will reduce the workload associated with periodic
renewal and installation of certificates on your servers. 
+ 
  These problems are compounded when organizations must use different domain names for each secure service
they make available on the Internet. In fact, organizations often find themselves in a situation where
they need to use sub-domain names -- names that use the same root name, but require a different prefix name
-- to secure each of the services they offer. Because prefix names are embedded into SSL certificates,
organizations usually buy one certificate per service. As you can imagine, this can become expensive and
time-consuming to manage, especially in organizations that run a multitude of Internet-facing
services. 
+ 
  Enter the multi-use SSL certificate. There are two types of multi-use certificates:
+ 
  Wildcard certificates can secure multiple sub-domains on a single unique Fully Qualified Domain Name
using a single certificate.
  Multi-domain certificates can secure multiple Fully Qualified Domain Names using a single certificate. 
  Each certificate can simplify management and reduce costs given the right situation.

  "'Wildcard Certificates'"
  The first type of multi-use certificate is the wildcard certificate. The name you embed in a certificate
must always follow the fully-qualified domain name (FQDN) format. If you want a certificate for the
Session Initiation Protocol (SIP) used by instant communication servers in the VirtualSpaceShip.com
domain, the name embedded into the certificate will be SIP.VirtualSpaceShip.com. If you want a
certificate for the email service, then you would normally have to buy a second certificate with the
second service name -- mail.VirtualSpaceShip.com -- embedded into it.
+ 
  In addition, some secure service implementations require internal as well as external validation and you
use a different name for each; for example, InternalSip.VirtualSpaceShip.com  and
ExternalSIP.VirtualSpaceShip.com. In this case, you must have a certificate on each server in the
internal and external service to allow users to work unimpeded whether they are in the office or on the
road. This is the case for instant messaging infrastructures where you want to ensure messages are
encrypted whether they are internal or external. Note that servers cannot include two certificates for
the same purpose.
+ 
  Wildcard certificates do not include service names. Instead, they are standard certificates that
support the use of a wildcard character to replace the prefix name in the subject name field, for example,
*.VirtualSpaceShip.com.  Using a wildcard certificate is much more practical and versatile than using
multiple single purpose certificate since the wildcard certificate can be applied to a number of
different services without requiring any updates. In addition, you can add, change or replace services
without needing to update the certificate. 

  For example, single wildcard certificate could easily support the following names and more:
www.VirtualSpaceShip.com, shop.VirtualSpaceShip.com, mail.VirtualSpaceShip.com,
SIP.VirtualSpaceShip.com, register.VirtualSpaceShip.com, and so on.
 <at>  <at>  -37, +45  <at>  <at> 

  The second type of multi-use certificate is the multi-domain certificate. While the wildcard
certificate will include a special character for the prefix name, the multi-domain name provides the
ability to include multiple Fully Qualified Domain Names within the same certificate.  However, unlike
wildcard certificates which can support an unlimited number of prefix names so long as the root domain
name remains the same, multi-domain certificates will only support the specific Fully Qualified Domain
Names entered into the certificate. In most cases, multi-domain certificates will support up to 25 or
more different Fully Qualified Domain Names in one certificate.

  Multi-domain certificates include the standard Subject Name field which supports a single primary
service name, as well as an additional entry called the Subject Alternative Name field which supports the
additional service names. The SAN certificate can therefore be installed on several servers and
function properly to support internal/external service delivery.
+ 
  SAN certificates have the same issues as single-purpose certificates however. Because the actual
service names are embedded into the certificate, you must make sure your services always use the same name
otherwise you must change the certificate and since the certificate is a multi-use certificate, you must
change it on each of the computers that host the service which the certificate supports. Additionally,
when you want to add services to provide further functionality to your users, you must update the SAN
certificate with the new service names.

  Multi-domain, SAN, or UCC certificates are useful when organizations require different root domain
names to run Internet-facing services.  Subject alternate name certificates are also called Unified
Communications Certificates (UCC) since they were primarily designed to support real-time
communications infrastructures. For example, an organization providing both internal and external
unified communications services with two different domain names -- for example,
SIP.VirtualSpaceShip.com the external domain and SIP.VirtualSpaceShip.com for the internal name --
would benefit from a multi-domain certificate because in this case, the wildcard certificate would not
work. In fact, if the organization was using wildcard certificates, two wildcard certificates would be
required because the root domain name is different in each case.
 <at>  <at>  -65, +74  <at>  <at> 

  Both certificate types offer reduced total cost of ownership (TCO) when they are deployed. But obviously,
both certificates only fit specific situations.

  Ideally, organizations would only use a single root name for all functions, but in most environments, this
is not possible. Many organizations use a least one public root name and one private root name to segregate
the internal from the external namespaces they work with. In this case, only multi-domain certificates
will work. But, if you only need a certificate for external purposes and you only use one single public root
name, then the wildcard certificate is the certificate of choice.
+ 
  In summary, multi-use certificates make it much easier to deploy multiple secure services both
internally and externally. This is particularly useful in environments that include several services
like mail, instant messaging, Web, mobile device management, and File Transfer Protocol (FTP). If this
is the case for your organization, then your best choice is to acquire the multi-use certificate that is
tailored to fit your needs. 

Apache Wiki | 9 Sep 2011 12:24
Picon
Favicon

[Httpd Wiki] Update of "ScratchPad" by jmcg

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "ScratchPad" page has been changed by jmcg:
http://wiki.apache.org/httpd/ScratchPad?action=diff&rev1=43&rev2=44

- Understanding Multi-Use Digital Certificates
+ = Understanding Multi-Use Digital Certificates =

- Abstract
+ == Abstract ==
  With the proliferation of internal Web-based services that must be exposed to the Internet,
organizations are turning more and more to the use of SSL certificates. However, using traditional SSL
certificates can become cumbersome and quite expensive, especially when organizations offer several
Internet-facing services. This is where multi-use certificates -- certificates that can be reused for
different purposes -- can help. Using a single multi-use certificate, organizations can reduce costs
and simplify certificate management.

- "'Introduction'"
+ == Introduction ==
  The best way to prove who you are on the Internet is to use a digital certificate. That's because a digital
certificate relies on a trusted, third-party authority to verify your identity. In fact, it uses a chain
of trust that begins with you and works its way up to the trusted authority that validates who you are. This
chain of trust provides verifiable Internet security.

  Take Internet user Bill for example. Bill uses a digital certificate to sign all of his emails. This example
demonstrates the concept of authenticity. The certificate authenticates Bill as the author of the email.
 <at>  <at>  -17, +17  <at>  <at> 

  
  Similarly, organizations that process transactions on the Internet or that offer Internet-based
services need to rely on digital certificates to validate that they are who they claim to be; otherwise, no
one will trust their services. Most organizations do this by adding certificates to their
Internet-facing servers. When users access a web page hosted on one of these servers, their Web browser
will automatically detect the certificate and modify the session, from an open session using the
HyperText Transfer Protocol (HTTP) to secure HTTP (HTTPS).   This will allow for the encryption of all the
data sent between the user's workstation and the server. HTTPS data encryption is provided by the Secure
Sockets Layer (SSL). Basically, SSL creates an encryption tunnel between the client and the server
protecting the transfer of data from one point to the other during the communication exchange. You know
you are using SSL when your browser displays a closed padlock in its status bar. 

- "'Working with SSL Certificates'"
+ == Working with SSL Certificates ==
  In order to enable SSL on their external-facing servers, organizations must purchase a certificate from a
trusted certification authority for each protected service they provide. Today, organizations can
offer several services to their end users -- email, instant messaging, mobile device management,
Web-based interactions and more -- and each one of the servers providing these services requires its own
certificate. 

  SSL certificates are tied to the unique domain name on which a service is hosted. Embedding the domain name
into the certificate is important since this makes it possible to ensure the identity of the remote
computer providing the service by comparing the domain name being accessed to the domain name included in
the certificate itself.
 <at>  <at>  -31, +31  <at>  <at> 

  Multi-domain certificates can secure multiple Fully Qualified Domain Names using a single certificate. 
  Each certificate can simplify management and reduce costs given the right situation.

- "'Wildcard Certificates'"
+ == Wildcard Certificates ==
  The first type of multi-use certificate is the wildcard certificate. The name you embed in a certificate
must always follow the fully-qualified domain name (FQDN) format. If you want a certificate for the
Session Initiation Protocol (SIP) used by instant communication servers in the VirtualSpaceShip.com
domain, the name embedded into the certificate will be SIP.VirtualSpaceShip.com. If you want a
certificate for the email service, then you would normally have to buy a second certificate with the
second service name -- mail.VirtualSpaceShip.com -- embedded into it.

  In addition, some secure service implementations require internal as well as external validation and you
use a different name for each; for example, InternalSip.VirtualSpaceShip.com  and
ExternalSIP.VirtualSpaceShip.com. In this case, you must have a certificate on each server in the
internal and external service to allow users to work unimpeded whether they are in the office or on the
road. This is the case for instant messaging infrastructures where you want to ensure messages are
encrypted whether they are internal or external. Note that servers cannot include two certificates for
the same purpose.
 <at>  <at>  -41, +41  <at>  <at> 

  For example, single wildcard certificate could easily support the following names and more:
www.VirtualSpaceShip.com, shop.VirtualSpaceShip.com, mail.VirtualSpaceShip.com,
SIP.VirtualSpaceShip.com, register.VirtualSpaceShip.com, and so on.
  In actual fact, the wildcard certificate supports the use of multiple sub-domain or prefix names within
the same certificate. Using a wildcard character as a placeholder in the domain name embedded into the
certificate makes the certificate much more versatile. In addition, it can be applied to any number of
uses since the wildcard character can represent any sub-domain name. Because of this, wildcard
certificates provide very good value for the cost.

- "'Multi-Domain Certificates'"
+ == Multi-Domain Certificates ==
  The second type of multi-use certificate is the multi-domain certificate. While the wildcard
certificate will include a special character for the prefix name, the multi-domain name provides the
ability to include multiple Fully Qualified Domain Names within the same certificate.  However, unlike
wildcard certificates which can support an unlimited number of prefix names so long as the root domain
name remains the same, multi-domain certificates will only support the specific Fully Qualified Domain
Names entered into the certificate. In most cases, multi-domain certificates will support up to 25 or
more different Fully Qualified Domain Names in one certificate.

  Multi-domain certificates include the standard Subject Name field which supports a single primary
service name, as well as an additional entry called the Subject Alternative Name field which supports the
additional service names. The SAN certificate can therefore be installed on several servers and
function properly to support internal/external service delivery.
 <at>  <at>  -55, +55  <at>  <at> 

  Multi-domain certificates are also useful for application service providers (ASP) who host
applications for multiple clients with each client using their own domain name. By using a multi-domain
certificate, ASPs can use a single certificate to support multiple clients. Note that the site seal and
certificate "Issued To" will only be for the primary domain name entered in the certificate and will not
include any of the other domain names. However, the certificate itself will include all of the domain
names that have been entered when the certificate was purchased. 

  While multi-domain certificates are also useful when used to support unified communications
deployments, there are some caveats for their use:
- - Multi-domain certificates do not support use of wildcard characters.  For this reason, sub-domain
names must be added as a unique domain name entries in the certificate. Each time a new domain name is added
or an old one is removed the certificate must be updated and re-deployed to each host server.
- - When hosting Web sites for multiple clients, ASPs should be aware that all domain names appear in a
multi-domain certificate. If the ASP does not want one site to seem as if it is connected to another, then a
different certificate type should be used.
-  Keep these caveats in mind when choosing a multi-domain certificate.

+  * Multi-domain certificates do not support use of wildcard characters.  For this reason, sub-domain
names must be added as a unique domain name entries in the certificate. Each time a new domain name is added
or an old one is removed the certificate must be updated and re-deployed to each host server.
+  * When hosting Web sites for multiple clients, ASPs should be aware that all domain names appear in a
multi-domain certificate. If the ASP does not want one site to seem as if it is connected to another, then a
different certificate type should be used.
+ 
+ Keep these caveats in mind when choosing a multi-domain certificate.
+ 
- "'Wildcard vs. Multi-Domain Certificates'"
+ == Wildcard vs. Multi-Domain Certificates ==
  Organizations wanting to move to subject alternate name or unified communication certificates should
choose between the wildcard and the multi-domain certificate types. Table 1 outlines the similarities
and differences between the two certificate formats.

  Both certificate types offer several benefits and include several features. In addition, both are
available in full authentication format only. Domain-only authentication certificates only require
the validation of the domain before they are issued, and because of this, cannot be used with multi-use
certificates. Full authentication certificates require both the validation of the domain itself and
the validation of the business running the domain. Because of this, full certificates are more
trustworthy than domain-only certificates. This is another reason the full authentication model is
used for multi-use certificates. Keep this in mind as you make your choice.

- "'Making the Selection'"
+ == Making the Selection ==
  Multi-use certificates were developed to provide multiple secure services originating from a single IP
address. To accomplish this, these certificates either add a subject alternate name (SAN) field to the
common single-use certificate or use a wildcard to replace the service name in the certificate.
Microsoft Exchange Server 2007 is an excellent example of this type of requirement. The same external
Exchange server can publish several different types of services: Outlook Web Access, Outlook Anywhere
Access, AutoDiscovery configuration information, and more. Each of these services requires the
publishing of its own name -- for example, OWA.VirtualSpaceShip.com, Mail.VirtualSpaceShip.com,
Autodiscover.VirtualSpaceShip.com --  ideally within a single certificate.

  Multi-use certificates reduce cost and simplify management by supporting the inclusion of multiple
names within the same certificate or the replacement of service names with a wildcard. However, each
multi-use certificate type should only be used in specific situations:

- * Wildcard certificates should be used when a single root name is used for all services or where there is a
single domain and only multiple sub-domains that cover all services. 
+  * Wildcard certificates should be used when a single root name is used for all services or where there is a
single domain and only multiple sub-domains that cover all services. 
- * Multi-domain certificates should be used when multiple root names are required for each service.
+  * Multi-domain certificates should be used when multiple root names are required for each service.
+ 
  Both certificate types offer reduced total cost of ownership (TCO) when they are deployed. But obviously,
both certificates only fit specific situations.

  Ideally, organizations would only use a single root name for all functions, but in most environments, this
is not possible. Many organizations use a least one public root name and one private root name to segregate
the internal from the external namespaces they work with. In this case, only multi-domain certificates
will work. But, if you only need a certificate for external purposes and you only use one single public root
name, then the wildcard certificate is the certificate of choice.
Apache Wiki | 9 Sep 2011 17:45
Picon
Favicon

[Httpd Wiki] Update of "CVE-2011-3192" by wrowe

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "CVE-2011-3192" page has been changed by wrowe:
http://wiki.apache.org/httpd/CVE-2011-3192?action=diff&rev1=1&rev2=2

Comment:
Draft edit from dirkx

- Describe CVE-2011-3192 here.
+ {{{
+           Apache HTTPD Security ADVISORY
+           ==============================
+                 UPDATE 3 - FINAL

+ Title:       Range header DoS vulnerability Apache HTTPD prior to 2.2.20.
+ 
+ CVE:         CVE-2011-3192
+ Last Change: 20110831 1800Z
+ Date:        20110824 1600Z
+ Product:     Apache HTTPD Web Server
+ Versions:    Apache 2.0 - all versions prior to 2.2.20;
+              Apache 1.3 is NOT vulnerable.
+ 
+ Changes since last update
+ =========================
+ 2.2.20 has a fix. 2.2.21 a bitter on. 1.3. not vulnerable. Further regex/rule
+ improments.  1.3 support stopgap module.  Explain DoS. Reduce severity for 1.3.
+ Added wiki link. Highlight fact that LimitRequestFieldSize is not sufficient.
+ 
+ Changes since update 1
+ =========================
+ In addition to the 'Range' header - the 'Request-Range' header is equally
+ affected. Furthermore various vendor updates, improved regexes (speed and
+ accommodating a different and new attack pattern).
+ 
+ Description:
+ ============
+ 
+ A denial of service vulnerability has been found in the way the multiple
+ overlapping ranges are handled by the Apache HTTPD server prior to version
+ 2.2.20:
+ 
+ An attack tool is circulating in the wild. Active use of this tool has
+ been observed.
+ 
+ The attack can be done remotely and with a modest number of requests can
+ cause very significant memory and CPU usage on the server.
+ 
+ The default Apache HTTPD installations version 2.0 and 2.2 prior to
+ 2.2.20 are vulnerable.
+ 
+ Apache 2.2.20 does fix this issue; however with a number of side effects
+ (see release notes). Version 2.2.21 xxx
+ 
+ Apache 1.3
+ ==========
+ 
+ Apache 1.3 is NOT vulnerable. However as explained in the background section
+ in more detail - this attack does cause a significant and possibly unexpected
+ load. You are advised to review your configuration in that light.
+ 
+ Type of Attack
+ ==============
+ 
+ This vulnerability concerns a 'Denial of Service' attack. This means that
+ a remote attacker, under the right circumstances, is able to slow your
+ service or server down to a craw. Leaving it unable to serve legitimate
+ clients in a timely manner.
+ 
+ There are no indications that this leads to a remote exploit; where a
+ third party can compromise your security and gain foothold of the server
+ itself. The result of this vulnerability is purely one of denying service
+ by grinding your server down to an halt.
+ 
+ Background and the 2007 report
+ ==============================
+ 
+ There are two aspects to this vulnerability. One is new, is Apache specific;
+ and resolved with this server side fix. The other issue is fundamentally a
+ protocol design issue dating back to 2007:
+ 
+ The contemporary interpretation of the HTTP protocol (currently) requires a
+ server to return multiple (overlapping) ranges; in the order requested. This
+ means that one can request a very large range (e.g. from byte 0- to the end)
+ 100's of times in a single request.
+ 
+ Being able to do so is an issue for (probably all) webservers and currently
+ subject of an IETF discussion to change the protocol:
+ 
+ This advisory details a problem with how Apache httpd and its so called
+ internal 'bucket brigades' deal with serving such "valid" request. The
+ problem is that currently such requests internally explode into 100's of
+ large fetches, all of which are kept in memory in an inefficient way. This
+ is being addressed in two ways. By making things more efficient. And by
+ weeding out or simplifying requests deemed too unwieldy.
+ 
+ FIX
+ ====
+ 
+ This vulnerability has been fixed in release 2.2.20  and beyond. You are
+ advised to upgrade to version 2.2.21 (or newer).
+ 
+ If you cannot upgrade - you can apply a Patch and recompile:
+ 
+ If you cannot upgrade and/or cannot apply above patches in a timely manner
+ then you could consider to apply te mitigations suggested below.
+ 
+ CAVEATS
+ =======
+ 
+ Note that this fix 1) will return a "200 OK" in cases where a 206 respond would
+ be larger and 2) changes the behavior of chunked responses. This may affect
+ certain clients. See the above background section and IETF reference for
+ more detail and the various discussions around fixing this in the protocol.
+ 
+ Furthermore a request with a byterange beyond the end of the file used to
+ return 416 but now returns 200. This is a violation of a RFC2616 SHOULD.
+ 
+ Mitigation:
+ ===========
+ 
+ There are several immediate options to mitigate this issue until a full fix
+ is available. Below examples handle both the 'Range' and the legacy
+ 'Request-Range' with various levels of care.
+ 
+ Note that 'Request-Range' is a legacy name dating back to Netscape Navigator
+ 2-3 and MSIE 3. Depending on your user community - it is likely that you
+ can use option '3' safely for this older 'Request-Range'.
+ 
+ 0) Consult
+  for more recent
+    information (as this is the final advisory).
+ 
+ 1) Use SetEnvIf or mod_rewrite to detect a large number of ranges and then
+    either ignore the Range: header or reject the request.
+ 
+    Option 1: (Apache 2.2)
+ 
+           # Drop the Range header when more than 5 ranges.
+           # CVE-2011-3192
+           SetEnvIf Range (?:,.*?){5,5} bad-range=1
+           RequestHeader unset Range env=bad-range
+ 
+           # We always drop Request-Range; as this is a legacy
+           # dating back to MSIE3 and Netscape 2 and 3.
+           #
+           RequestHeader unset Request-Range
+ 
+           # optional logging.
+           CustomLog logs/range-CVE-2011-3192.log common env=bad-range
+ 
+    Above may not work for all configurations. In particular situations
+    mod_cache and (language) modules may act before the 'unset'
+    is executed upon during the 'fixup' phase.
+ 
+    Option 2: (Pre 2.2)
+ 
+           # Reject request when more than 5 ranges in the Range: header.
+           # CVE-2011-3192
+           #
+           RewriteEngine on
+           RewriteCond %{
+ } !(^bytes=[^,]+(,[^,]+){0,4}$|^$) [NC]
+           RewriteRule .* - [F]
+ 
+           # We always drop Request-Range; as this is a legacy
+           # dating back to MSIE3 and Netscape 2 and 3.
+           #
+           RequestHeader unset Request-Range
+ 
+    The number 5 is arbitrary. Several 10's should not be an issue and may be
+    required for sites which for example serve PDFs to very high end eReaders
+    or use things such complex http based video streaming.
+ 
+    WARNING These directives need to be specified in every configured
+    vhost, or inherited from server context as described in:
+ 
+ 2) Use mod_headers to completely dis-allow the use of Range headers:
+ 
+           RequestHeader unset Range
+ 
+    Note that this may break certain clients - such as those used for
+    e-Readers and progressive/http-streaming video.
+ 
+    Furthermore to ignore the Netscape Navigator 2-3 and MSIE 3 specific
+    legacy header - add:
+ 
+           RequestHeader unset Request-Range
+ 
+    Unlike the commonly used 'Range' header - dropping the 'Request-Range'
+    is not likely to affect many clients.
+ 
+ 4) Deploy a Range header count module as a temporary stopgap measure.
+ 
+    An improved stop-gap module for the 2.x series was provided by
+    Guenter Knauf and can be found at:
+ 
+ Note
+ ====
+ 
+ Earlier advisories suggested theuse of LimitRequestFieldSize. This method is
+ not fully effective and can by bypassed by splitting the attack vector up
+ across multiple headers. Therefore you should not rely on LimitRequestFieldSize
+ alone.
+ 
+ OS and Vendor specific information
+ ==================================
+ 
+ Red Hat:        Has additional RHEL specific information at:
+ 
+ NetWare:        Pre compiled binaries available.
+ 
+ mod_security:   Has updated their rule set; see
+ 
+ 
+ Actions:
+ ========
+ 
+ Apache HTTPD users who are concerned about a DoS attack against their server
+ should 1) upgrade to version 2.2.20, 2) if not possible - apply the provided
+ patches or 3) consider implementing any of the above mitigations immediately.
+ 
+ When using a third party attack tool to verify vulnerability - note that most
+ of the versions in the wild currently check for the presence of mod_deflate;
+ and will (mis)report that your server is not vulnerable if this module is not
+ present. This vulnerability is not dependent on presence or absence of
+ that module.
+ 
+ Planning:
+ =========
+ 
+ No further advisories are planned. However we will track information at
+ }}}
+ 
Apache Wiki | 9 Sep 2011 17:52
Picon
Favicon

[Httpd Wiki] Update of "CVE-2011-3192" by wrowe

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "CVE-2011-3192" page has been changed by wrowe:
http://wiki.apache.org/httpd/CVE-2011-3192?action=diff&rev1=2&rev2=3

  
  Changes since last update
  =========================
- 2.2.20 has a fix. 2.2.21 a bitter on. 1.3. not vulnerable. Further regex/rule
- improments.  1.3 support stopgap module.  Explain DoS. Reduce severity for 1.3.
+ 2.2.20 has a fix, 2.2.21 an improved one. Version 1.3 is not vulnerable. 
+ Further regex/rule improvements.  Explained DoS.  Added wiki link.  
- Added wiki link. Highlight fact that LimitRequestFieldSize is not sufficient.
+ Highlight fact that LimitRequestFieldSize workaround was insufficient.

  Changes since update 1
  =========================
 <at>  <at>  -120, +120  <at>  <at> 

  2-3 and MSIE 3. Depending on your user community - it is likely that you
  can use option '3' safely for this older 'Request-Range'.

+ 0) Consult http://httpd.apache.org/security/CVE-2011-3192.txt for the most
- 0) Consult
-  for more recent
-    information (as this is the final advisory).
+    recent information (as this is the final advisory).

  1) Use SetEnvIf or mod_rewrite to detect a large number of ranges and then
     either ignore the Range: header or reject the request.
 <at>  <at>  -191, +190  <at>  <at> 

  Note
  ====

- Earlier advisories suggested theuse of LimitRequestFieldSize. This method is
+ Earlier advisories suggested the use of LimitRequestFieldSize. This mitigation
- not fully effective and can by bypassed by splitting the attack vector up
+ was not fully effective and can by bypassed by splitting the attack vector up
  across multiple headers. Therefore you should not rely on LimitRequestFieldSize
  alone.

 <at>  <at>  -210, +209  <at>  <at> 

  ========

  Apache HTTPD users who are concerned about a DoS attack against their server
- should 1) upgrade to version 2.2.20, 2) if not possible - apply the provided
- patches or 3) consider implementing any of the above mitigations immediately.
+ should 1) upgrade to version 2.2.21 (or 2.0.65 when it becomes available), 
+ 2) if not possible - apply the provided patches or 3) consider implementing 
+ any of the above mitigations immediately.

  When using a third party attack tool to verify vulnerability - note that most
  of the versions in the wild currently check for the presence of mod_deflate;
 <at>  <at>  -222, +222  <at>  <at> 

  Planning:
  =========

- No further advisories are planned. However we will track information at
+ No further advisory email announcements are planned. However we will track 
+ minor refinements of this advisory at;
+ 
+   http://httpd.apache.org/security/CVE-2011-3192.txt
+ 
+ Further recommendations and discussion on workarounds, or user-agent
+ specific complications of these fixes will be tracked at;
+ 
+   http://wiki.apache.org/httpd/CVE-2011-3192
+ 
  }}}

Apache Wiki | 9 Sep 2011 17:59
Picon
Favicon

[Httpd Wiki] Update of "CVE-2011-3192" by wrowe

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "CVE-2011-3192" page has been changed by wrowe:
http://wiki.apache.org/httpd/CVE-2011-3192?action=diff&rev1=3&rev2=4

  overlapping ranges are handled by the Apache HTTPD server prior to version
  2.2.20:

+      http://seclists.org/fulldisclosure/2011/Aug/175 
+ 
  An attack tool is circulating in the wild. Active use of this tool has
  been observed.

 <at>  <at>  -72, +74  <at>  <at> 

  and resolved with this server side fix. The other issue is fundamentally a
  protocol design issue dating back to 2007:

+       http://seclists.org/bugtraq/2007/Jan/83 
+ 
  The contemporary interpretation of the HTTP protocol (currently) requires a
  server to return multiple (overlapping) ranges; in the order requested. This
  means that one can request a very large range (e.g. from byte 0- to the end)
 <at>  <at>  -79, +83  <at>  <at> 

  
  Being able to do so is an issue for (probably all) webservers and currently
  subject of an IETF discussion to change the protocol:
+ 
+       http://trac.tools.ietf.org/wg/httpbis/trac/ticket/311

  This advisory details a problem with how Apache httpd and its so called
  internal 'bucket brigades' deal with serving such "valid" request. The
 <at>  <at>  -91, +97  <at>  <at> 

  ====

  This vulnerability has been fixed in release 2.2.20  and beyond. You are
- advised to upgrade to version 2.2.21 (or newer).
+ advised to upgrade to version 2.2.21 (or newer, or 2.0.65 once that version
+ is published).

  If you cannot upgrade - you can apply a Patch and recompile:
+ 
+   http://www.apache.org/dist/httpd/patches/apply_to_2.2.14/ (for 2.2.9 - .14)
+   http://www.apache.org/dist/httpd/patches/apply_to_2.2.19/ (for 2.2.15 - .19)
+   http://www.apache.org/dist/httpd/patches/apply_to_2.0.64/ (for 2.0.55 - .64)

  If you cannot upgrade and/or cannot apply above patches in a timely manner
  then you could consider to apply te mitigations suggested below.
 <at>  <at>  -152, +163  <at>  <at> 

            #
            RewriteEngine on
            RewriteCond %{
+ HTTP:range
  } !(^bytes=[^,]+(,[^,]+){0,4}$|^$) [NC]
            RewriteRule .* - [F]

Apache Wiki | 9 Sep 2011 18:00
Picon
Favicon

[Httpd Wiki] Update of "CVE-2011-3192" by wrowe

Dear Wiki user,

You have subscribed to a wiki page or wiki category on "Httpd Wiki" for change notification.

The "CVE-2011-3192" page has been changed by wrowe:
http://wiki.apache.org/httpd/CVE-2011-3192?action=diff&rev1=4&rev2=5

  
  Changes since last update
  =========================
- 2.2.20 has a fix, 2.2.21 an improved one. Version 1.3 is not vulnerable. 
+ 2.2.20 has a fix, 2.2.21 an improved one. Version 1.3 is not vulnerable.
- Further regex/rule improvements.  Explained DoS.  Added wiki link.  
+ Further regex/rule improvements.  Explained DoS.  Added wiki link.
  Highlight fact that LimitRequestFieldSize workaround was insufficient.

  Changes since update 1
 <at>  <at>  -33, +33  <at>  <at> 

  overlapping ranges are handled by the Apache HTTPD server prior to version
  2.2.20:

-      http://seclists.org/fulldisclosure/2011/Aug/175 
+      http://seclists.org/fulldisclosure/2011/Aug/175

  An attack tool is circulating in the wild. Active use of this tool has
  been observed.
 <at>  <at>  -74, +74  <at>  <at> 

  and resolved with this server side fix. The other issue is fundamentally a
  protocol design issue dating back to 2007:

-       http://seclists.org/bugtraq/2007/Jan/83 
+       http://seclists.org/bugtraq/2007/Jan/83

  The contemporary interpretation of the HTTP protocol (currently) requires a
  server to return multiple (overlapping) ranges; in the order requested. This
 <at>  <at>  -162, +162  <at>  <at> 

            # CVE-2011-3192
            #
            RewriteEngine on
+           RewriteCond %{HTTP:range} !(^bytes=[^,]+(,[^,]+){0,4}$|^$) [NC]
-           RewriteCond %{
- HTTP:range
- } !(^bytes=[^,]+(,[^,]+){0,4}$|^$) [NC]
            RewriteRule .* - [F]

            # We always drop Request-Range; as this is a legacy
 <at>  <at>  -221, +219  <at>  <at> 

  ========

  Apache HTTPD users who are concerned about a DoS attack against their server
- should 1) upgrade to version 2.2.21 (or 2.0.65 when it becomes available), 
+ should 1) upgrade to version 2.2.21 (or 2.0.65 when it becomes available),
- 2) if not possible - apply the provided patches or 3) consider implementing 
+ 2) if not possible - apply the provided patches or 3) consider implementing
  any of the above mitigations immediately.

  When using a third party attack tool to verify vulnerability - note that most
 <at>  <at>  -234, +232  <at>  <at> 

  Planning:
  =========

- No further advisory email announcements are planned. However we will track 
+ No further advisory email announcements are planned. However we will track
  minor refinements of this advisory at;

    http://httpd.apache.org/security/CVE-2011-3192.txt
 <at>  <at>  -243, +241  <at>  <at> 

  specific complications of these fixes will be tracked at;

    http://wiki.apache.org/httpd/CVE-2011-3192
- 
  }}}


Gmane