has anyone successfully added IIS script maps to their Wix installation. I have been using this:
<CustomAction Id="AddHandlers_Cmd" Property="AddHandlers" Value=""[SystemFolder]inetsrv\appcmd" set config -section:system.webServer/handlers /+"[\[]name='GSOAP',path='*.dll', verb='*',modules='IsapiModule',scriptProcessor='[BIN]mod_gsoap.dll',resourceType='File', requireAccess='Execute'[\]]"" Execute="immediate"/>
<CustomAction Id="AddHandlers" BinaryKey="WixCA" DllEntry="CAQuietExec" Execute="deferred" Return="check" Impersonate="no"/>
However I get the 404 errors from my web application and when I add the script map manually after installing the application it works perfectly. I would appreciate if anyone could share how they done this.
It sounds like you need to add handler mapping to your web application. If that's the case, then you don't need a custom action to accomplish this.
Take a look at WebApplicationExtension element.
We are writing a separate Custom action in C# and Schedule it in after install Finalize to add handlers.
Try this, you can do anything with your own custom action.
I haven't worked with GSOAP handler, but used ASP.NET ISAPI extention. I suppose basically it has the same level of integration. I will post the code for ASP.NET and I hope you won't have problem with changing it to GSOAP. I have a huge risk of changing something incorrectly and make the examples unusable for you, so I let you do it better.
Here is my code to register extention itself if it isn't
<Component Id="Iis6ConfigExtentions" Guid="{GuidHash({ProductId}-1822E4F3-5850-47D5-9281-D1E0E20C77D4)}" KeyPath="yes" Permanent="yes">
<Condition><![CDATA[Installed OR (IISMAJORVERSION AND (IISMAJORVERSION = "#6"))]]></Condition>
<iis:WebServiceExtension Id="ExtensionASP4" Group="ASP.NET v4.0.30319" Allow="yes" File="[FRAMEWORKROOT]v4.0.30319\aspnet_isapi.dll" Description="ASP.NET v4.0.30319"/>
</Component>
and for adding those to the concrete site you need to use WebApplicationExtention, as #Yan already mentioned:
<iis:WebApplication Id="IIS6WebApp" Name="[SITE_APP_NAME]" WebAppPool="AppPool" >
<iis:WebApplicationExtension Verbs="GET,HEAD,POST" CheckPath="no" Script="yes" Executable="[FRAMEWORKROOT]v4.0.30319\aspnet_isapi.dll" />
</iis:WebApplication>
The CheckPath="no" is required on IIS 6 because we use ASP.NET MVC which doesn't map urls to files on disk. You might need it too since service address probably doesn't map to physical file.
UPDATE:
Looks like it is really not possible to do it directly in WIX (at least in 3.5 version). You are going in right direction - custom action using appcmd. But the provided custom actions aren't complete or what you need: the first one just sets property value to the command you need to execute and the second one calls some Dll custom action which wasn't provided.
What I'd try to execute is:
<CustomAction Id="AddGsoapHandler" Directory="TARGETDIR" Impersonate="no" Execute="deferred" Return="check" ExeCommand="[SystemFolder]inetsrv\appcmd.exe set config -section:system.webServer/handlers /+"[\[]name='GSOAP',path='*.dll', verb='*',modules='IsapiModule',scriptProcessor='[BIN]mod_gsoap.dll',resourceType='File', requireAccess='Execute'[\]]" /commit:apphost" />
Please pay attention to the path attribute of your handler configuration. It will limit handler usage to only URL's which ends with ".dll" - I just copied it from your code but not sure if you've done it intentionally or just copied from example at Microsoft site. That might have led to 401 error even if handler is registered successfully.
Also I added the /commit:apphost argument. Not sure if you really need this, but I use it to explicitly specify that I need to configure my application's config.
UPDATE 2:
If you need to set configuration to only single application, you should specify application object as argument of appcmd instead of the 'config' which means global configuration. Please try this:
<CustomAction Id="AddGsoapHandler" Directory="TARGETDIR" Impersonate="no" Execute="deferred" Return="check" ExeCommand="[SystemFolder]inetsrv\appcmd.exe set app "MySite/app1" -section:system.webServer/handlers /+"[\[]name='GSOAP',path='*.dll', verb='*',modules='IsapiModule',scriptProcessor='[BIN]mod_gsoap.dll',resourceType='File', requireAccess='Execute'[\]]" /commit:apphost" />
For more details you can read this tutorial.
Related
This seems like a relatively simple question, but I didn't see it clearly indicated in the documentation I examined. Assuming a typical Elmah usage on a Windows Server 2012/IIS 7 webserver installation, if the Elmah web.config settings are pointed to an Elmah.config file, will altering that file trigger under the running application cause IIS to restart?
While the answer provided by Josh would normally be true, it is not in the case of ELMAH. ELMAH instantiates a new error log implementation every time it is needed, reading the configuration from either web.config or the configSource provided.
You can do a simple test to prove this. If you configure ELMAH to use an external config file for the errorLog element like this:
<elmah>
<errorLog configSource="elmah.config"/>
</elmah>
and your elmah.config looks like this:
<?xml version="1.0" encoding="utf-8"?>
<errorLog type="Elmah.XmlFileErrorLog, Elmah" logPath="~/App_Data" />
When you start the project and force an error, a new XML file is written to the App_Data folder. Change the logPath like this (without restarting the application):
<?xml version="1.0" encoding="utf-8"?>
<errorLog type="Elmah.XmlFileErrorLog, Elmah" logPath="~/App_Data2" />
(I changed to App_Data2).
Force a new error and you will now see ELMAH create the App_Data2 folder on the file system.
No: Straight from MSDN:
Saving a Web.config file restarts the application. You can alternatively use the configSource attribute of individual section elements to point to a secondary configuration file that does not cause an application restart when it is changed. For more information, see configSource in General Attributes Inherited by Section Elements.
MSDN
I wrote an interceptor in one of my projects to intercept all the requests. So usually in a spring project i will do normal build and start the server and my changes related to an interceptor will start reflecting. However this doesn't seem to be the case with a hybris project.
Do I need to do update in hybris hac as well? And if I do then out of the available below mentioned options which options do i need to choose and why.
1. Update running system
2. Clear the hMC configuration from the database
3. Create essential data
4. Localize types
Thanks,
Ashish
To answer the second part of your question, I have listed at least one reason for selecting each type of options. Hope this helps.
Update running system - Required when hybris type definition changes. For example, changing the content of file <extnesion-name>-items.xml
Clear the hMC configuration from the database - If you have chosen to persist hmc configuration in the Database and changing it. For example, changing content in the file hmc.xml
Create essential data - If there is a change in the content in the impex files which follows the naming pattern of essentialdata*.impex.
Localize types - If there is a change in the properties files for localizations. for example changing the content in file <extension-name>-locales_en.properties
Assuming you've not made any changes to any underlying data objects (Jalo items) then you won't need to run an update in the hybris hAC.
You should just be able to make your changes, run ant clean all from the platform and then start up the hybris ECP instance and your changes will be visible.
If this is a normal Spring MVC interceptor, then it should work fine.
Have you checked your spring configuration in the storefront extension you are working from?
For example, take a look at the accelerator Spring MVC configuration file:
hybris/bin/ext-template/yacceleratorstorefront/web/webroot/WEB-INF/config/spring-mvc-config.xml
This has some examples where this is used out-of-the-box:
<mvc:interceptors>
<ref bean="beforeControllerHandlerInterceptor" />
<ref bean="beforeViewHandlerInterceptor" />
<ref bean="csrfHandlerInterceptor" />
</mvc:interceptors>
As an example, the default before controller handler interceptor is defined as:
<!-- Interceptor that runs once per request and before the controller handler method is called -->
<alias name="defaultBeforeControllerHandlerInterceptor" alias="beforeControllerHandlerInterceptor" />
<bean id="defaultBeforeControllerHandlerInterceptor" class="de.hybris.platform.yacceleratorstorefront.interceptors.BeforeControllerHandlerInterceptor" >
<property name="beforeControllerHandlers">
<ref bean="beforeControllerHandlersList" />
</property>
</bean>
which references:
<alias name="defaultBeforeControllerHandlersList" alias="beforeControllerHandlersList" />
<util:list id="defaultBeforeControllerHandlersList" >
<!-- List of handlers to run -->
<bean class="de.hybris.platform.yacceleratorstorefront.interceptors.beforecontroller.SecurityUserCheckBeforeControllerHandler" />
<bean class="de.hybris.platform.yacceleratorstorefront.interceptors.beforecontroller.RequireHardLoginBeforeControllerHandler" >
<property name="userService" ref="userService"/>
<property name="redirectStrategy" ref="redirectStrategy"/>
...
</bean>
<bean class="de.hybris.platform.yacceleratorstorefront.interceptors.beforecontroller.DeviceDetectionBeforeControllerHandler" />
...
</util:list>
So you could either override this using the alias with your own implementation, or add additional controller handlers to the list.
As there is no change to the underlying data model - this is just wiring up Spring MVC related classes - no need for an update system or anything like that. Just an 'ant clean all' to recompile to pick up your new interceptor classes, and server restart to pick up the change in the Spring cornfiguraton.
I am using IIS 7.5 on Windows 2008 Server R2 and have installed the IIS Rewrite Module v2.0. I am trying to enable logging to see exactly what rewrite rule is being used. Based on research I have learned that I need to modify (update) C:\Windows\System32\inetsrv\config\applicationHost.config. In order to do so I have created a NEW XML schema file called "RewriteLogging.xml" and have placed it in C:\Windows\System32\inetsrv\config\schema directory. The file contains the following code:
<!--
IIS Rewrite Logging
-->
<configSchema>
<sectionSchema name="system.applicationHost/RewriteLogging">
<attribute name="RewriteLog" type="string" defaultValue="C:\inetpub\logs\LogFiles\rewrite.txt" />
<attribute name="RewriteLogLevel" type="int" defaultValue="9" />
</sectionSchema>
</configSchema>
After restarting the web server I notice that the applicationHost.config file's date and time stamp get updated but my new directives are not getting added.
So here are my questions:
Am I approaching this the correct way?
Am I correct in creating a separate XML file and placing it in the schema directory?
Should I be creating a NEW section Schema name or should I be using an existing one? If so which one?
Am I using the correct syntax in the XML file? When I open it up in a browser I am not seeing any errors which is encouraging.
Am I missing something altogether or forgetting an important step?
Any help would be GREATLY appreciated.
I just started checking out Windows Azure and I have trouble getting any access logs from IIS for my test web role. The web role itself works fine, but I would like to see a log for accesses (both successful and failed).
As far as I can see the default configuration files for a web role contain instructions to send those logs to a blob named "wad-iis-logfiles", but that blob is never even created (it doesn't exist in my blob storage).
My diagnostics.wadcfg for the web role currently is:
<DiagnosticMonitorConfiguration configurationChangePollInterval="PT1M" overallQuotaInMB="4096" xmlns="http://schemas.microsoft.com/ServiceHosting/2010/10/DiagnosticsConfiguration">
<DiagnosticInfrastructureLogs bufferQuotaInMB="512" scheduledTransferPeriod="PT5M" />
<Directories bufferQuotaInMB="512" scheduledTransferPeriod="PT5M">
<IISLogs container="wad-iis-logfiles" />
<CrashDumps container="wad-crash-dumps" />
</Directories>
<Logs bufferQuotaInMB="512" scheduledTransferPeriod="PT5M" scheduledTransferLogLevelFilter="Information" />
<PerformanceCounters bufferQuotaInMB="512">
(... snip...)
</PerformanceCounters>
<WindowsEventLog bufferQuotaInMB="512" scheduledTransferPeriod="PT1M" scheduledTransferLogLevelFilter="Error">
<DataSource name="Application!*" />
</WindowsEventLog>
</DiagnosticMonitorConfiguration>
Question 1: is this configuration file correct?
Question 2: are there other things that need to be set before I can get the IIS log files?
With the help of the commenters I was able to solve the issue.
There are several interacting things causing the issue.
As commenter #kwill mentioned, an existing configuration blob in wad-control-container overrides any other configuration, and that configuration is not replaced during an in-place update. I was using in-place update to put my modified diagnostics.wadcfg in place, so that is the explanation why my attempts to change settings that way didn't work. Note that editing the properties of the web test role (in the "Roles" branch of the Azure Cloud services project operates by editing that same file, so that didn't work either. More information on how that wad-control-container overrides setting can be found in http://msdn.microsoft.com/en-us/library/windowsazure/dn205146.aspx .
The reason that blob already existed may have been that I had been changing some other performance measurement settings in the azure management window earlier.
I managed to "fix" the situation by editing the blob found in wad-control-container for my instance, using the tool mentioned by commenter #Gaurav Mantri - "Azure Explorer". As mentioned, without that tool you can download the blob and edit it, but never put it back properly, since the '/' characters in the blob's name get translated to '%2F', and those are not translated back on upload.
Note that the XML schema is not the same as the schema for diagnostics.wadcfg, but some similarities exist. I changed the "Directories" element toward the bottom of the blob to read:
<Directories>
<BufferQuotaInMB>512</BufferQuotaInMB>
<ScheduledTransferPeriodInMinutes>2</ScheduledTransferPeriodInMinutes>
<Subscriptions>
<DirectoryConfiguration>
<Path>C:\Resources\directory\8091b0be14e54213ac12fcbd5f9c8e1b.WebTestRole.DiagnosticStore\CrashDumps</Path>
<Container>wad-crash-dumps</Container>
<DirectoryQuotaInMB>0</DirectoryQuotaInMB>
</DirectoryConfiguration>
<DirectoryConfiguration>
<Path>C:\Resources\directory\8091b0be14e54213ac12fcbd5f9c8e1b.WebTestRole.DiagnosticStore\LogFiles</Path>
<Container>wad-iis-logfiles</Container>
<DirectoryQuotaInMB>16</DirectoryQuotaInMB>
</DirectoryConfiguration>
</Subscriptions>
</Directories>
In the original version the "BufferQuotaInMB" and "DirectoryQuotaInMB" fields were 0.
Note that after uploading the blob again the effect is not immediate. It takes a while for the changed configuration to get picked up, and then it takes another while before the IIS log files are copied for the first time.
Last note: it may be obvious, but I don't think editing that blob is a recommendable solution. It is good to know the option exists though.
Is it possible to use appcmd to change the value of allowDefinition? Specifically I'm try to enable changes to the httpCompression module at the application level.
Modifying the applicationHost.config by manually changing the following line:
<section name="httpCompression" allowDefinition="AppHostOnly" overrideModeDefault="Deny" />
To
<section name="httpCompression" allowDefinition="MachineToApplication" overrideModeDefault="Allow" />
allows me to then execute the following appcmd commands:
appcmd set config "website name" /section:httpCompression /noCompressionForProxies:false
appcmd set config "website name" /section:httpCompression /noCompressionForHttp10:false
However I need a solution that does not rely on manually editing the applicationHost.config
My problem was, I had to change anonymousAuthentication from False to True. When I did:
appcmd set config websitename /section:anonymousAuthentication /enabled:True
Error I got:
Config Error This configuration cannot be used at this path. This happens when the section is locked at the parent level. Locking is either by default(overrideModeDefault="Deny")...
To unlock, do the following:
appcmd unlock config /section:? This will list the section you want. Then type:
appcmd unlock config /section:system.webserver/security/authentication/anonymousauthentication
Thats it... :)
Try using %windir%\system32\inetsrv\appcmd unlock config -section:*sectionName*. See http://blog.donnfelker.com/2007/03/26/iis-7-this-configuration-section-cannot-be-used-at-this-path/
I actually came across a need to do just that after posting this answer.
%systemroot%\System32\inetsrv\appcmd.exe unlock config /section:system.WebServer/[rest of the path to config section you need to edit]
One big warning, you should NEVER change the allowDefinition, that is an important setting that is usually there for a reason, for example it might be that even if you set it in a specific directory or app it will not work, so the developers have specified that.
So please, never modify the allowDefinition attribute in the section definitions. On the other hand you can modify the overrideModeDefault which will allow users to define it in a different place if allowed by definition.