Search Results

Search found 7960 results on 319 pages for 'distributed cache'.

Page 22/319 | < Previous Page | 18 19 20 21 22 23 24 25 26 27 28 29  | Next Page >

  • How to tell nginx to honor backend's cache?

    - by ChocoDeveloper
    I'm using php-fpm with nginx as http server (I don't know much about reverse proxies, I just installed it and didn't touch anything), without Apache nor Varnish. I need nginx to understand and honor the http headers I send. I tried with this config (taken from the docs) but didn't work: /etc/nginx/nginx.conf: fastcgi_cache_path /var/lib/nginx/cache levels=1:2 keys_zone=website:10m inactive=10m; fastcgi_cache_key "$scheme$request_method$host$request_uri"; /etc/nginx/sites-available/website: server { fastcgi_cache website; #fastcgi_cache_valid 200 302 1h; #fastcgi_cache_valid 301 1d; #fastcgi_cache_valid any 1m; #fastcgi_cache_min_uses 1; #fastcgi_cache_use_stale error timeout invalid_header http_503; add_header X-Cache $upstream_cache_status; } I always get "MISS" and the cache dir is empty. If I uncomment the other directives, I get hit, but I don't want those "dumb" settings, I need to control them within my backend. For example, if my backend says "public, s-maxage=10", the cache should be considered stale after 10 secs. Instead, nginx will store it for 1h, because of these directives. I was thinking whether I should try proxy_cache, not sure what's the difference. In both fastcgi and proxy modules docs it says this: The cache honors backend's Cache-Control, Expires, and etc. since version 0.7.48, Cache-Control: private and no-store only since 0.7.66, though. Vary handling is not implemented. nginx version: nginx/1.1.19 Any thoughts? pd: I also have the reverse proxy that is offered by Symfony2 (which I turn off to use nginx's). The headers are interpreted correctly by it, so I think I'm doing it right.

    Read the article

  • Why is apt-cache so slow?

    - by Damn Terminal
    After upgrade to Trusty (14.04) from Saucy (13.10), all apt operations are very slow. Even those that do not include downloading anything, or connecting to any servers. For example, displaying the apt policy # time apt-cache policy [...] real 0m8.951s user 0m5.069s sys 0m3.861s takes almost ten seconds! Mostly a weird lag right after issuing the command. And it's the same even if I issue the same command again. On another system it doesn't take a tenth of a second real 0m0.096s user 0m0.070s sys 0m0.023s The other system is a little beefier but there was no noticeable difference before the upgrade. It's the same with apt-get, anything apt-related. How do I find out the source of this lag and fix it? Additional info: # cat /etc/nsswitch.conf # /etc/nsswitch.conf # # Example configuration of GNU Name Service Switch functionality. # If you have the `glibc-doc-reference' and `info' packages installed, try: # `info libc "Name Service Switch"' for information about this file. passwd: compat group: compat shadow: compat hosts: files dns networks: files protocols: db files services: db files ethers: db files rpc: db files netgroup: nis BTW is my understanding of how apt-cache works correct? It doesn't make any network connections when I run apt-cache policy, right? In case I'm wrong and it matters, here are my sources https://gist.github.com/anonymous/02920270ff68e23fc3ec

    Read the article

  • Building a Distributed Commerce Infrastructure in the Cloud using Azure and Commerce Server

    - by Lewis Benge
    One of the biggest questions I routinely get asked is how scalable Commerce Server is. Of course the text book answer is the product has been around for 10 years, powers some of the largest e-Commerce websites in the world, so it scales horizontally extremely well. One argument however though is what if you can't predict the growth of demand required of your Commerce Platform, or need the ability to scale up during busy seasons such as Christmas for a retail environment but are hesitant on maintaining the infrastructure on a year-round basis? The obvious answer is to utilise the many elasticated cloud infrastructure providers that are establishing themselves in the ever-growing market, the problem however is Commerce Server is still product which has a legacy tightly coupled dependency on Windows and IIS components. Commerce Server 2009 codename "R2" however introduced to the concept of an n-tier deployment of Microsoft Commerce Server, meaning you are no longer tied to core objects API but instead have serializable Commerce Entity objects, and business logic allowing for Commerce Server to now be built into a WCF-based SOA architecture. Presentation layers no-longer now need to remain on the same physical machine as the application server, meaning you can now build the user experience into multiple-technologies and host them in multiple places – leveraging the transport benefits that a WCF service may bring, such as message queuing, security, and multiple end-points. All of this logic will still need to remain in your internal infrastructure, for two reasons. Firstly cloud based computing infrastructure does not support PCI security requirements, and secondly even though many of the legacy Commerce Server dependencies have been abstracted away within this version of the application, it is still not a fully supported to be deployed exclusively into the cloud. If you do wish to benefit from the scalability of the cloud however, you can still achieve a great Commerce Server and Azure setup by utilising both the Azure App Fabric in terms of the service bus, and authentication services and Windows Azure to host any online presence you may require. The architecture would be something similar to this: This setup would allow you to construct your Commerce Services as part of your on-site infrastructure. These services would contain all of the channels custom business logic, and provide the overall interface back into the underlying Commerce Server components. It would be recommended that services are constructed around the specific business domain of the application, which based on your business model would usually consist of separate services around Catalogue, Orders, Search, Profiles, and Marketing. The App Fabric service bus is then used to abstract and aggregate further the services, making them available to the cloud and subsequently secured by App Fabrics authentication services. These services are now available for consumption by any client, using any supported technology – not just .NET. Thus meaning you are now able to construct apps for IPhone, integrate with Java based POS Devices, and any many other potential uses. This aggregation is useful, and forms the basis of the further strategy around diversifying and enhancing the e-Commerce experience, but also provides the foundation for the scalability we want to gain from utilising a cloud-based application platform. The Windows Azure application platform is Microsoft solution to benefiting from the true economies of scale in terms of the elasticity of the cloud. Just before the launch of the Azure Platform – Domino's pizza actually managed to run their whole SuperBowl operation from the scalability of Windows Azure, and simply switching back to their traditional operation the next day with no residual infrastructure costs. The platform also natively can subscribe to services and messages exposed within the AppFabric service bus, making it an ideal solution to build and deploy a presentation layer which will need to support of scalable infrastructure – such as a high demand public facing e-Commerce portal, or a promotion element of a brand. Windows Azure has excellent support for ASP.NET, including its own caching providers meaning expensive operations such as catalogue queries can persist in memory on the application server, reducing the demand on internal infrastructure and prioritising it for more business critical operations such as receiving orders and processing payments. Windows Azure also supports other languages too, meaning utilising this approach you can technically build a Commerce Server presentation layer in Java, PHP, or Ruby – or equally in ASP.NET or Silverlight without having to change any of the underlying business or Commerce Server implementation. This SOA-style architecture is one of the primary differentiators for Commerce Server as a product in the e-Commerce market, and now with the introduction of a WCF capability in Commerce Server 2009/2009 R2 the opportunities for extensibility of the both the user experience, and integration into third parties, are drastically increased, all with no effect to the underlying channel logic. So if you are looking at deployment options for your e-Commerce application to help support demand in a cost effective way. I would highly recommend you consider looking at Windows Azure, and if you have any questions in-particular about this style of deployment, please feel free to get in touch!

    Read the article

  • Distributed transactions and queues, ruby, erlang

    - by chrispanda
    I have a problem that involves several machines, message queues, and transactions. So for example a user clicks on a web page, the click sends a message to another machine which adds a payment to the user's account. There may be many thousands of clicks per second. All aspects of the transaction should be fault tolerant. I've never had to deal with anything like this before, but a bit of reading suggests this is a well known problem. So to my questions. Am I correct in assuming that secure way of doing this is with a two phase commit, but the protocol is blocking and so I won't get the required performance? It appears that DBs like redis and message queuing system like Rescue, RabbitMQ etc don't really help me a lot - even if I implement some sort of two phase commit, the data will be lost if redis crashes because it is essentially memory-only. All of this has led me to look at erlang - but before I wade in and start learning a new language, I would really like to understand better if this is worth the effort. Specifically, am I right in thinking that because of its parallel processing capabilities, erlang is a better choice for implementing a blocking protocol like two phase commit, or am I confused?

    Read the article

  • Distributed C++ game server which use database.

    - by Slav
    Hello. My C++ turn-based game server (which uses database) does stand against current average amount of clients (players), so I want to expand it to multiple (more then one) amount of computers and databases where all clients still will remain within single game world (servers will must communicate with each other and use multiple databases). Is there some tutorials/books/common standards which explain how to do it in a best way?

    Read the article

  • Entity Framework and distributed Systems

    - by Dirk Beckmann
    I need some help or maybe only a hint for the right direction. I've got a system that is sperated into two applications. An existing VB.NET desktop client using Entity Framework 5 with code first approach and a asp.net Web Api client in C# that will be refactored right yet. It should be possible to deliver OData. The system and the datamodel is still involving and so migrations will happen in undefined intervalls. So I'm now struggling how to manage my database access on the web api system. So my favourd approch would be us Entity Framework on both systems but I'm running into trouble while creating new migrations. Two solutions I've thought about: Shared Data Access dll The first idea was to separate the data access layer to a seperate project an reference from each of the systems. The context would be the same as long as the dll is up to date in each system. This way both soulutions would be able to make a migration. The main problem ist that it is much more complicate to update a web api system than it is with the client Click Once Update Solution and not every migration is important for the web api. This would couse more update trouble and out of sync libraries Database First on Web Api The second idea was just to use the database first approch an on web api side. But it seems that all annotations will be lost by each model update. Other solutions with stored procedures have been discarded because of missing OData support and maintainability. Does anyone run into same conflicts or has any advices how such a problem can be solved!

    Read the article

  • How are Reads Distributed in a Workload

    - by Bill Graziano
    People have uploaded nearly one millions rows of trace data to TraceTune.  That’s enough data to start to look at the results in aggregate.  The first thing I want to look at is logical reads.  This is the easiest metric to identify and fix. When you upload a trace, I rank each statement based on the total number of logical reads.  I also calculate each statement’s percentage of the total logical reads.  I do the same thing for CPU, duration and logical writes.  When you view a statement you can see all the details like this: This single statement consumed 61.4% of the total logical reads on the system while we were tracing it.  I also wanted to see the distribution of reads across statements.  That graph looks like this: On average, the highest ranked statement consumed just under 50% of the reads on the system.  When I tune a system, I’m usually starting in one of two modes: this “piece” is slow or the whole system is slow.  If a given piece (screen, report, query, etc.) is slow you can usually find the specific statements behind it and tune it.  You can make that individual piece faster but you may not affect the whole system. When you’re trying to speed up an entire server you need to identity those queries that are using the most disk resources in aggregate.  Fixing those will make them faster and it will leave more disk throughput for the rest of the queries. Here are some of the things I’ve learned querying this data: The highest ranked query averages just under 50% of the total reads on the system. The top 3 ranked queries average 73% of the total reads on the system. The top 10 ranked queries average 91% of the total reads on the system. Remember these are averages across all the traces that have been uploaded.  And I’m guessing that people mainly upload traces where there are performance problems so your mileage may vary. I also learned that slow queries aren’t the problem.  Before I wrote ClearTrace I used to identify queries by filtering on high logical reads using Profiler.  That picked out individual queries but those rarely ran often enough to put a large load on the system. If you look at the execution count by rank you’d see that the highest ranked queries also have the highest execution counts.  The graph would look very similar to the one above but flatter.  These queries don’t look that bad individually but run so often that they hog the disk capacity. The take away from all this is that you really should be tuning the top 10 queries if you want to make your system faster.  Tuning individually slow queries will help those specific queries but won’t have much impact on the system as a whole.

    Read the article

  • Scrum Board for a distributed team

    - by Falcon
    I am looking for recommendations on a digital Scrum Board which can be shared over the internet. I imagine something like a big tablet on which you can draw and which remote users can access, too. I dislike Scrum software because I think one major benefit of a Scrum Board is its physical presence. It should be hard to ignore. The best solution would be two big tablets on which you can draw and which can be synchronized. Has anyone got product recommendations for something like this? Or would you rather use a software? Kind regards, Falcon

    Read the article

  • Headaches using distributed version control for traditional teams?

    - by J Cooper
    Though I use and like DVCS for my personal projects, and can totally see how it makes managing contributions to your project from others easier (e.g. your typical Github scenario), it seems like for a "traditional" team there could be some problems over the centralized approach employed by solutions like TFS, Perforce, etc. (By "traditional" I mean a team of developers in an office working on one project that no one person "owns", with potentially everyone touching the same code.) A couple of these problems I've foreseen on my own, but please chime in with other considerations. In a traditional system, when you try to check your change in to the server, if someone else has previously checked in a conflicting change then you are forced to merge before you can check yours in. In the DVCS model, each developer checks in their changes locally and at some point pushes to some other repo. That repo then has a branch of that file that 2 people changed. It seems that now someone must be put in charge of dealing with that situation. A designated person on the team might not have sufficient knowledge of the entire codebase to be able to handle merging all conflicts. So now an extra step has been added where someone has to approach one of those developers, tell him to pull and do the merge and then push again (or you have to build an infrastructure that automates that task). Furthermore, since DVCS tends to make working locally so convenient, it is probable that developers could accumulate a few changes in their local repos before pushing, making such conflicts more common and more complicated. Obviously if everyone on the team only works on different areas of the code, this isn't an issue. But I'm curious about the case where everyone is working on the same code. It seems like the centralized model forces conflicts to be dealt with quickly and frequently, minimizing the need to do large, painful merges or have anyone "police" the main repo. So for those of you who do use a DVCS with your team in your office, how do you handle such cases? Do you find your daily (or more likely, weekly) workflow affected negatively? Are there any other considerations I should be aware of before recommending a DVCS at my workplace?

    Read the article

  • Page cache flushing behavior under heavy append load

    - by Bryce
    I'm trying to understand the behavior of the Linux pdflush daemon when: The page cache is initially pretty much empty There is a large amount of free memory The system starts undergoing heavy write load My understanding right now is that the vm.dirty_ratio and vm.dirty_background_ratio that control page cache flushing behavior are with respect to the present size of the page cache, which means that my writes will flush earlier than they would if the page cache was pre-populated (even with dummy data from some random file), and thus throughput will be lower. Is this accurate?

    Read the article

  • Why does Tomcat try to use the cache when compilation failed?

    - by etheros
    For some reason, it appears Tomcat is trying to hit its compilation cache when compilation failed. For example, if I create a JSP containing nothing but Hello, <%=world%>!, predictably, I get an error: org.apache.jasper.JasperException: Unable to compile class for JSP. Subsequent requests however alternate between this and org.apache.jasper.JasperException: org.apache.jasper.JasperException: Unable to load class for JSP. Further, if I create a JSP containing Hello!, it of course works just fine. If I modify it contain Hello, <%=name%>!, the response alternates between the previously-mentioned compilation error, and the cached Hello!. What's going on?

    Read the article

  • Does ZFS cache Compressed or Uncompressed data in a ZFS file-system with compression turned on?

    - by George Bailey
    ZFS supports file-system compression and it also caches frequently or recently accessed data. If a system has lots of CPU but the underlying data storage system is slow. It is possible that ZFS would perform better with compression turned on. This can be easily tested when writing files by measuring CPU and disk usage and throughput. (of course latency may exist,, but this would not be an issue for large files). But what about cache? If data will have to be decompressed every time it is read then this is probably less of a good idea. Is the cached data compressed?. Does anybody have some information on this?

    Read the article

  • How can I diagnose cache misses when using Apache as a reverse proxy?

    - by johnstok
    I have set up Apache 2.2 as a reverse proxy with the following configuration: # jBoss proxying ProxyRequests Off <Proxy *> Order deny,allow Allow from all </Proxy> ProxyPass /foo http://localhost:9080/foo ProxyPassReverse /foo http://localhost:9080/foo ProxyPassReverseCookiePath /foo /foo # Reverse proxy caching CacheEnable disk /foo # Compression SetOutputFilter DEFLATE BrowserMatch ^Mozilla/4 gzip-only-text/html BrowserMatch ^Mozilla/4\.0[678] no-gzip BrowserMatch \bMSIE\s(7|8) !no-gzip !gzip-only-text/html DeflateCompressionLevel 9 Header append Vary User-Agent env=!dont-vary However, in a number of cases where I expect a cached response to be returned the request is sent through to the origin server at localhost:9080. Responses have a HTTP Vary header of 'Accept-Encoding,User-Agent' which is to be expected given the mod_deflate configuration. How can I determine why Apache is unable to serve a response from the cache?

    Read the article

  • Any way to get back Chrome's Dialog box for cache clearing instead of the new tab?

    - by Stuart P.
    As of today's release of chrome (Tuesday, March 8, 2011) on both Mac & PC the settings are now in a tab (chrome://settings/advanced), needless to say when you're clearing your cache very frequently (cmd-shift-delete on mac, cntl+shift+delete on PC) it's quite tedious going back and forth in tabs. The click & clean chrome extension doesn't have a mac counterpart (plus I like the keyboard much more than the mouse). I've searched and have yet to find a way to get a dialog box instead of the new tab.

    Read the article

  • Tell the linux kernel to put a file in the disk cache?

    - by Rory
    Is there any command to for a file to be read in and loaded into the linux disk cache? This is on an up-to-date debian system. I know in the general case, it's better to let the linux kernel figure this out. But I have an edge case. I have a laptop that has an NFS director mounted, and i want to play a long video file, but I don't want to have a network problem interrupt the playnig. I know that (largeish) file will be read in it's entirety later on. I know that nothing else (really) will be running while playing this video. There is enough free memory to store this file. (I know I could just copy the file into a new tmpfs filesystem, but I'm curious if there's an even shorter way to do it)

    Read the article

  • ubuntu is very slow

    - by johnny smithens
    Hello all. I am new with Ubuntu, and it is very slow(even in Ubuntu 2D). The performance is degraded for almost any task. I just reinstalled with amd64, and tried updating the Nvidia drivers with Nvidia Xserver. but it made no difference. This is the output of free -m: total used free shared buffers cached Mem: 3006 1318 1688 0 61 699 -/+ buffers/cache: 556 2449 Swap: 3064 0 3064 tl;dr - total: 3006, used: 1318 When I see the virtual console with Ctrl+Alt+F2, I see constantly: Assuming Drive Cache: write through; asking for cache data failed; It is very frustrating. Thanks in advance!

    Read the article

< Previous Page | 18 19 20 21 22 23 24 25 26 27 28 29  | Next Page >