Search results matching tag 'disk cache'

Re: how to use the ARR for Image Protect

wonyoo — Fri, 09 Oct 2009 18:17:59 GMT

The anti-leeching can be done using URL rewrite (which is installed with ARR).

tip #6

http://ruslany.net/2009/04/10-url-rewriting-tips-and-tricks/

ARR v2 RC Disk Cache Drive Marked Unhealthy - EventID's 1006 & 1007 vs. 1008

JAltrichter — Fri, 09 Oct 2009 11:37:39 GMT

Since installing the ARR v2 RC I am periodically receiving warning events that the ARR disk cache directory is unhealthy.

"Drive with path '\\?\path' is being marked unhealthy. The data contains the error code."

Provider: Application Request Routing
EventID: 1006 [ Qualifiers] 32768
Event Data: 7B 00 07 80

About 45 to 90 seconds later there is a corresponding information message (EventID 1007) that the drive is being marked healthy again.

Prior to installing the ARR v2 RC (using the ARR v2 Beta 2 release) I did not see this message pattern. Instead the server would record the following information message:

EventID 1008 - "Scavenger successfully ran for content in disk location"

Is this an anomaly with RC release?

Jim

Re: ARR2 caching and sizing

wonyoo — Fri, 25 Sep 2009 03:28:25 GMT

It's difficult to make a blanket recommendation on CPU/memory without knowing the traffic pattern of your usage. As you may have guessed, the first bottleneck that you are going to hit is disk i/o. Depending on the average size of the files that are being cached, you may or may not be able to saturate the NIC.

That said, we do have some deployment recommendations that are applicable to ARRv2 deployment in general. It can be found at Deployment recommendations for Application Request Routing.

As for your last question on DVR on live smooth streaming (LSS), it depends on what you mean by that. In an upcoming release of LSS, there is a role that gets enabled when ARR is installed.

Re: ARR v2 RC Memory Cache Duration Change

wonyoo — Thu, 24 Sep 2009 23:54:14 GMT

Jim

You are absolutely right. We did change the default setting from 5 to 60 seconds.

We made this change with the CDN customers in mind where 5 second caching was simply too short (ie. Publicly available Internet contents that are cache-able don't really change every 5 seconds. And by keeping it at 5, ARR will have to check for the "freshness" unnecessarily too frequently.)

In our testing that simulates CDN, we determined that 60 was more appropriate default value that balances between performance and resource consumption.

ARR Disk Cache Drive Health Status Unknown

JAltrichter — Thu, 24 Sep 2009 21:03:27 GMT

I have just installed the ARR v2 RC and wanted to ask a questions that I did not bother with during the beta.

In the Application Request Routing Cache form the health status of my primary drive appears as Unknown with an Alert of "Disk Utilization is not available because the worker process is not running."

What service do I need to enable to properly query the health status of a disk cache drive?

I was reading back through the examples on configuring ARR and noticed that even in your updated examples (http://learn.iis.net/page.aspx/575/configure-and-enable-disk-cache-in-application-request-routing/) the drive health status is shown as Unknown.

Thanks in advance,

Jim

ARR v2 Caching and ASP.Net Works Great, Why?

JAltrichter — Sun, 30 Aug 2009 23:16:57 GMT

I would like to understand how ARR disk caching works with ASP.Net applications and have not been able to find a good reference.

I have a simple testing server farm configuration with a single ARR server and a few content servers. Everything with my ASP.Net application is caching appropriately, static images and CSS sheets are cached and dynamic data is excluded. I do not understand how this is working.

I have browsed through the web.config file for the application, but I am just not seeing it. Can you give me a gentle nudge in the right direction.

Thanks in advance.

Jim

Re: ARR v2 Beta 2 Disk Cache Statistics

wonyoo — Tue, 25 Aug 2009 20:23:27 GMT

With memory (and in conjuction with disk), it works slightly differently. Let me see if I can clarify.

Let's suppose that the memory cache interval is set to 5 seconds (default value.)

The flow would be something like:

1) Request is received by Windows server (http.sys, actually).

2) Cache miss at the kernel

3) Request forwarded to IIS/ARR

4) Cache miss at the disk

5) Request forwarded to origin and content received by IIS/ARR

6) Content cached to disk

7) Content cached to memory (for 5 seconds)

(Note that this is a two tiered caching system - and not counting the extreme corner cases, what's cached in memory will almost always be a subset of what's cached on disk.)

At T + 2 seconds

8) Same request received by Windows

9) Cache hit at memory (by http.sys)

10) response sent to the client

At T + 6 seconds (greater than the configured 5 second value)

11) SAme request received by Windows

12) Cache miss at memory (for memory caching, it will unload the object when it doesn't need it. This is one of the core differences between memory caching and disk caching. Memory is more scarce than disk - so after 5 (or configured) seconds, they are unloaded. They may be unloaded more aggressively depending on memory consumption, etc but that's all managed by kernel.)

13) Request forwarded to IIS/ARR

14) Disk cache hit (and it will also check for the freshness)

15) Content cached once again in memory (for 5 seconds - or configured value).

Hope that explains.

Re: ARR v2 Beta 2 Disk Cache Statistics

JAltrichter — Tue, 25 Aug 2009 15:23:48 GMT

Won,

Thanks that was great.

To carry over an idea from our other thread (http://forums.iis.net/t/1160295.aspx) I assume that irrespective of whether the object is cached in memory or disk, the object invalidation process with be the same. The cached object will always be served to the requesting user first then checked for accuracy as a secondary task.

Jim

Re: ARR Disk Cache Invalidation Process

JAltrichter — Tue, 25 Aug 2009 15:13:22 GMT

I assume the prescribed solution for this condition, might be create a Global URL rewrite rule on the ARR server that redirects input for a blank URL explicitly to the default page.

If that rule processes before the ARR Cache Control rule then I would think you would get the same behaviour for the default page as all other HTML pages.

Jim

Re: ARR v2 Beta 2 Disk Cache Statistics

wonyoo — Tue, 25 Aug 2009 15:09:55 GMT

Jim

The memory cache is done at the kernel level. You could have a netmon between ARR and the content server to see if you actually see the cache hit.

A simpler way would be to type the following on an elevated command prompt (which will show you what's stored in kernel cache.)

netsh http show cache

If the object is "cacheable", then the Windows kernel will cache the content. ARR leverages this feature and changes the duration when appropriate. That said, if you have a content cached for 5 seconds and changed it to 50, the change won't impact what's already cached. It has to be invalidated (ie. the 5 seconds must expire) before the change (50 seconds) takes effect with the next "cache-miss" request.

Now, if you see your content in the responses from above command, you can trust us that it's being served from memory :).

Does that help?