Search Results

Search found 2454 results on 99 pages for 'domains'.

Page 23/99 | < Previous Page | 19 20 21 22 23 24 25 26 27 28 29 30  | Next Page >

• Using my own Postfix, filtering spam and getting all the mail into my ISP's inbox

  - by djechelon

  Hello, I currently own a domain bought via GoDaddy.com, which provides me a basic email setup for the most common needs. I configured it to forward all mail to [email protected] to [email protected]. I also own a virtual server with a running Postfix that I use for a specific website (all mail to somedomain.com gets forwarded via LMTP to a program written by me). Since I'm recently experiencing some harassing by spammers, since GoDaddy doesn't seem to filter spam, and since my Windows Phone's Pocket Outlook cannot filter spam, I would like to use SpamAssassin to filter inbound spam by changing my domain's MX records to my server My ideal setup is the following: All mail delivered to somedomain.com gets redirected via LMTP as usual via virtual transport without any spam check All mail to [email protected] gets redirected to [email protected] after a severe spam check I don't care about [email protected] since I use just one address for now I would like to train SpamAssassin with customized spam rules, possibly based on the presence of certain keywords (links to certain unsubscribe pages I found recurring) I currently configured Postfix with transport somedomain.com lmtp:[127.0.0.1]:8025 .somedomain.com error: Cannot accept mail for this domain relay somedomain.com OK (I guess I should add mydomain.com OK too) virtual @mydomain.com [email protected] (looks like a catch-all rule, it's OK as requirement 3) I installed SpamAssassin, I can do rcspamd start and set it to boot with the server, but I don't know if there is anything else to do for use in Postfix, and how to apply requirement 1 (only mail to mydomain.com gets filtered) I also tried to send an email via Telnet to make sure my settings are ready for MX change. I received the message into my account but I found that it gone through secureserver.net, like Postfix didn't rewrite the destination but simply relayed the message. Thank you in advance. I'm no expert in SpamAssassin, and I have little experience in Postfix (enough to avoid making my server an open relay)

  Read the article

• Splitting Company into two separate companies with duplicate IT infrastructure. Need Reccomendation

  - by Shanghai360

  We are dividing our company into two companies. All employees will be shared across both the companies. We have separate Accounting systems, email and other applications for both companies. There are two physical offices located within the same city block. And identical infrastructure at both. Money is not a limiting factor. How would you design the network, remote access, and configure the workstations? Thanks!

  Read the article

• How to deploy new instances of the same application (on 1 server) automatically?

  - by Intru

  I'm working on a SaaS application where each customer runs its own version of the application. All the application instances currently run on a single server. This works quite well for us (we need less resources in total). The application doesn't use a lot of resources, so even a small VPS would be overkill (and more expensive). Adding a new customer is currently quite a bit of work: Create a user that is allowed to ssh Create a new MySQL database and user Create a virtual host for the application Log in with the new user, do a git checkout of the application (in the right location) Create tables in the new database, and add some init data Add some cron jobs Create a first user that can log in Add this new instance to capistrano What would be the best way to automate these tasks? Are the applications that can (given proper configuration) do this? Ideally this should be usable for a sales-person (so something web-based). I could write a (bash) script that does most of these tasks, and then maybe add a small web-based wrapper where someone could provider the domain/default user information. Of course, this would also require a delete-script, since some customers will eventually leave, which means that you need a list of all existing customers/instances.

  Read the article

• Iis multiple domain binding

  - by user2979484

  I've a server with vshpere running and one vm with Windows 2008 r2. I'm hosting a website on this vm. But I want to try to add another domain. Now I have 3 ip-adresses. 1 for vspehere. 1 for vm. 1 for 2nd domain. When I add a new site and edit the bindings to the last ip-adress (2nd domain) it doesn't work. I've added also a new network adapter on vsphere and edited in the windows network adapter. Could anyone help me out with this. Sorry for my bad english I hope you will understand it.

  Read the article

• One Apache server, multiple clients - best practices for config files?

  - by OttaSean

  First time user; please be gentle. :-) (And if you don't like my question I'd be grateful for a comment as to why...) I am doing a contract at a government server shop that provides web services for multiple client groups in other areas of the government. My employer has asked me to look into how other shops, in similar situations, handle configuration files, and whether there are any best practices on the subject. I'm pretty sure there are lots of installations out there running multiple VirtualHosts out of one Apache installation, but surprisingly I couldn't find anything online about how people handle config file layout, so was hoping some of you wise folks on ServerFault might have some thoughts or pointers for me. The current setup - which seems logical to me - is that each client site has its own directory off the root - so: /client/tps-reports/ /client/silly-walks/ /client/ministry-of-magic/ and so on - and each of those directories has a /htdocs, /cgi-bin, and /conf (among others). The main /etc/apache/httpd.conf only contains Include statements (and lots of comments), the last of which is: Include /etc/apache/vhosts/*.conf The vhosts directory contains symlinks: tpsrept.conf - /client/tps-reports/conf/tpsrept.conf sillywk.conf - /client/silly-walks/conf/sillywk.conf mom.conf - /client/ministry-of-magic/mom.conf Each of those .conf files contains the actual NameVirtualHost definition and a gigantic <VirtualHost 192.168.12.34> stanza - which contains all the stuff about the specific site. The idea is that clients have access to what's in their own /client/xx directory, so they can change stuff in the section of the config that is relevant to them. As I mentioned above, that seems fairly logical to me, but I'm wondering if any of you wise folks are aware of potential gotchas with this sort of layout, or any other thoughts on why it is or isn't a good idea. In particular, how do other places do it? Is there a "best practice" for this sort of thing? Many thanks in advance for your time and any thoughts you all might have.

  Read the article

• WCF service using duplex channel in different domains

  - by ds1

  I have a WCF service and a Windows client. They communicate via a Duplex WCF channel which when I run from within a single network domain runs fine, but when I put the server on a separate network domain I get the following message in the WCF server trace... The message with to 'net.tcp://abc:8731/ActiveAreaService/mex/mex' cannot be processed at the receiver, due to an AddressFilter mismatch at the EndpointDispatcher. Check that the sender and receiver's EndpointAddresses agree. So, it looks like the communication just work in one direction (from client to server) if the components are in two separate domains. The Network domains are fully trusted, so I'm a little confused as to what else could cause this? Server app.config <?xml version="1.0" encoding="utf-8" ?> <configuration> <system.serviceModel> <behaviors> <serviceBehaviors> <behavior name="JobController.ActiveAreaBehavior"> <serviceMetadata httpGetEnabled="false" /> <serviceDebug includeExceptionDetailInFaults="true" /> </behavior> </serviceBehaviors> </behaviors> <services> <service behaviorConfiguration="JobController.ActiveAreaBehavior" name="JobController.ActiveAreaServer"> <endpoint address="mex" binding="mexTcpBinding" bindingConfiguration="" contract="IMetadataExchange" /> <host> <baseAddresses> <add baseAddress="net.tcp://SERVER:8731/ActiveAreaService/" /> </baseAddresses> </host> </service> </services> </system.serviceModel> </configuration> but I also add an end point programmatically in Visual C++ host = gcnew ServiceHost(ActiveAreaServer::typeid); NetTcpBinding^ binding = gcnew NetTcpBinding(); binding->MaxBufferSize = Int32::MaxValue; binding->MaxReceivedMessageSize = Int32::MaxValue; binding->ReceiveTimeout = TimeSpan::MaxValue; binding->Security->Mode = SecurityMode::Transport; binding->Security->Transport->ClientCredentialType = TcpClientCredentialType::Windows; ServiceEndpoint^ ep = host->AddServiceEndpoint(IActiveAreaServer::typeid, binding, String::Empty); // Use the base address Client app.config <?xml version="1.0" encoding="utf-8" ?> <configuration> <system.serviceModel> <bindings> <netTcpBinding> <binding name="NetTcpBinding_IActiveAreaServer" closeTimeout="00:01:00" openTimeout="00:01:00" receiveTimeout="00:10:00" sendTimeout="00:01:00" transactionFlow="false" transferMode="Buffered" transactionProtocol="OleTransactions" hostNameComparisonMode="StrongWildcard" listenBacklog="10" maxBufferPoolSize="524288" maxBufferSize="65536" maxConnections="10" maxReceivedMessageSize="65536"> <readerQuotas maxDepth="32" maxStringContentLength="8192" maxArrayLength="16384" maxBytesPerRead="4096" maxNameTableCharCount="16384" /> <reliableSession ordered="true" inactivityTimeout="00:10:00" enabled="false" /> <security mode="Transport"> <transport clientCredentialType="Windows" protectionLevel="EncryptAndSign" /> <message clientCredentialType="Windows" /> </security> </binding> </netTcpBinding> </bindings> <client> <endpoint address="net.tcp://SERVER:8731/ActiveAreaService/" binding="netTcpBinding" bindingConfiguration="NetTcpBinding_IActiveAreaServer" contract="ActiveArea.IActiveAreaServer" name="NetTcpBinding_IActiveAreaServer"> <identity> <userPrincipalName value="[email protected]" /> </identity> </endpoint> </client> </system.serviceModel> </configuration> Any help is appreciated! Cheers

  Read the article

• Visual studio 2010 remote debugging is very slow (across domains, over vpn)

  - by alex

  Overall debugging works, but each step through code takes dosens of seconds. I've already closed all additional windows like stack trace,watches,autos; deleted all breakpoints. server and dev machine are located in different domains, so i set up local user on both, with matching password. remote debugger is running as service. looking at security log, I found quite a lot of entries about remote debugging account logging in (record about every minute). Any suggestions on how i can speed up remote debugging? dev computer: quad core, 8 Gb mem, win 7 x64 , visual studio 2010 ultimate target server: asp.net website , 2xdual core xeon, 2Gb mem, remote debugger 2010 communication channel: vpn , 5 mbit , latency about 20ms. (seems that debbugging never uses more than 20 kb/s)

  Read the article

• How to Test a Multi-Tenant App with support for multiple domains

  - by asifch

  HI, we are building a multi-tenant application, which will support that each tenant can have a unique top level domain, the application is build using the asp.net 3.5 and SQL servr 2005, while each tenant will have different database. I have seen a number of questions about the similar applications on the StackOverFlow, but none of them is related to the Testing, I want to know is how one can test the application in a development environment, specially How can we test that each customer connects to his own DB based on the URL. how can we emulate different domains on the local system. like abc.com and xyz.com all goes to dev machine's IIS. Any recommendations which might help us in the developing process of such an application.

  Read the article

• Dot Net Nuke - Sub Domains

  - by Tristan

  Hi Guys, I've got a DNN portal set up for me and my brother, which works, and i'd like to make it so that sub domains point to our respective sub section of the sites. I.e. brothers.co.uk = brothers.co.uk/home.aspx me.brothers.co.uk = brothers.co.uk/me.aspx him.brothers.co.uk = brothers.co.uk/him.aspx I've got the sub sections set up in DNN, but can't figure out what to do with domain names to get it working. I'd like this to be all one site with the same login set, although if there's a better way to structure this, I'm open to ideas. Any thoughts? Regards Tristan

  Read the article

• Jabber integration into site and multiple domains

  - by Jamie

  Hi all, I'm creating a centralised web product, i.e. a customer gets assigned a personal domain, for example company1.example.com or company2.example.com and can then use our service. I'm planning to integrate a jabber service into the website. I have already found a decent jabber client library which I can use for the site. I know that jabberd2 uses mysql, which is perfect because I want to use the web interface to add users, delete users, view message logs etc. However, my problem is when I have two companies or more. I would like a jabber server which can host multiple domains, i.e. jabber.company1.example.com, jabber.company2.example.com Do you have any experience of this? And do you know of a good jabber server which will do this? Any help would be hugely appreciated!

  Read the article

• Rails - Multiple top level domains and a single session/cookie

  - by Thadius B

  Hello all, I've been struggling with this for quite awhile and haven't been able to find a solution. I need a user to be able to view multiple top level domains with a single login. My understanding is that this needs to be set in environment.rb and called with before_dispatch. This is what I've come up with: require 'activesupport' require 'dispatcher' module ActionController class Dispatcher def set_session_domain ActionController::Base.session_options.update :session_domain => "#{@request.host}" end before_dispatch :set_session_domain end end However, this does not seem to be working when I try and pull the values from session[:session_domain]. Any help is greatly appreciated.

  Read the article

• Apache mod_rewrite and multiple domains

  - by andihahn

  Hi, I'm trying to use mod_rewrite to map multiple domains to different servlets on one host. Example: www.dom1.com - 192.168.1.n/dom1 www.dom2.com - 192.168.1.n/dom2 ... I'm using the mod_rewrite and mod_proxy and directive but it seems that the reverse mapping via ProxyPassReverse doesn't work as I expected. ProxyPassReverse /subdomain.domain.com http://192.168.1.n/subdomain doesn't work. I've turned rewrite-logging on with RewriteLog /var/log/rewrite.log From the logs I'd say that rewriting works and the problem seems to be with reverse mapping. However I can't see any Reverse mapping entries. It seems that reverse mapping isn't logged or needs a different command to be activated. (Apache and the servlet container are on different machines but this should not matter I'd think ?)

  Read the article

• How to get a list of all domains?

  - by AngryHacker

  I'm trying to get all domains that are available in the Windows Login dialog (in the Domain dropdown). I've tried the following code but it only returns the domain I am logged into. Am I missing something? StringCollection domainList = new StringCollection(); try { DirectoryEntry en = new DirectoryEntry("LDAP://"); // Search for objectCategory type "Domain" DirectorySearcher srch = new DirectorySearcher("objectCategory=Domain"); SearchResultCollection coll = srch.FindAll(); // Enumerate over each returned domain. foreach (SearchResult rs in coll) { ResultPropertyCollection resultPropColl = rs.Properties; foreach( object domainName in resultPropColl["name"]) { domainList.Add(domainName.ToString()); } } } catch (Exception ex) { Trace.Write(ex.Message); } return domainList;

  Read the article

• Hide other domains' menus from node edit form on a Drupal site using domain access

  - by monkeyninja

  I'm in the process of making some improvements to a live Drupal site that's using the Domain Access module to run a number of microsites. I'm trying to find a way of restricting the menus a user can post content to from the node edit screen. A user on one of the domains should only be able to post content to menus associated with that domain. Is there a simple way of achieving this? I'm guessing there are some hooks I could use, but so far I have been unable to identify them. I'd prefer not to have to install further modules to achieve this and to be able to add some code to the current site to alter the forms. The site is struggling with the large number of modules we've had to install on it already.

  Read the article

• Rails - Dynamic cookie domains using Rack

  - by Tim B.

  I'm fairly new to Rails and Rack, but this guy had a seemingly straightforward write-up about using Rack to implement dynamic session domain middleware. The code looks good to and I've implemented it here on my local machine, but I'm still not able to transcend top level domains on a single login. Here's the middleware code: class SetCookieDomain def initialize(app, default_domain) @app = app @default_domain = default_domain end def call(env) host = env["HTTP_HOST"].split(':').first env["rack.session.options"][:domain] = custom_domain?(host) ? ".#{host}" : "#{@default_domain}" @app.call(env) end def custom_domain?(host) domain = @default_domain.sub(/^\./, '') host !~ Regexp.new("#{domain}$", Regexp::IGNORECASE) end end And then in environment.db: config.load_paths += %W(#{RAILS_ROOT}/app/middlewares) Lastly in production.db (and development.db): config.middleware.use "SetCookieDomain", ".example.org" Any help is greatly appreciated. EDIT: I'm running Rails 2.3.3 and Rack 1.0

  Read the article

• The Difference Between .com, .net, .org and Why We’re About To See Many More Top-Level Domains

  - by Chris Hoffman

  .com, .net, .org and other website suffixes are known as “top-level domains” (TLDs). While we normally see only a few of these, there are hundreds of them – and there may be thousands more soon. Top-level domains are managed by the Internet Assigned Numbers Authority (IANA), which is run by the Internet Corporation for Assigned Names and Numbers (ICANN). HTG Explains: What is the Windows Page File and Should You Disable It? How To Get a Better Wireless Signal and Reduce Wireless Network Interference How To Troubleshoot Internet Connection Problems

  Read the article

• Does the SPDY protocol eliminate the need for cookieless domains?

  - by Clint Pachl

  With plain HTTP, cookieless domains were an optimization to avoid unnecessarily sending cookie headers for page resources. However, the SPDY protocol compresses HTTP headers and in some cases eliminates unnecessary headers. My question then is, does SPDY make cookieless domains irrelevant? Furthermore, should the page source and all of its resources be hosted at the same domain in order to optimize a SPDY implementation?

  Read the article

• Access SSAS cube from across domains without direct database connection

  - by SuperKing

  Hello, I'm working with SQL Server Analysis Services for the first time and have the dilemma of working on a project in which users must be able to access SSAS Cubes (via a custom web dashboard) that live across different servers and domains, but without having access to the other server's SSAS database connection strings. So Organization A and Organization B will have their own cubes on their own servers, but Organization A users must be able to view Organization B's cubes, and Organization B users must be able to view Organization A's cubes, but neither organization should have access to the connection string. I've read about allowing HTTP access to the SSAS server and cube from the link below, but that requires setting up users for authentication or allowing anonymous access to one organization's server for users of another organization, and I'm not sure this would be acceptable for this situation, or if this is the preferred way to do this. Is performance acceptable here? http://technet.microsoft.com/en-us/library/cc917711.aspx I also wonder if perhaps it makes sense to run a nightly/weekly process that accesses the other organization's SSAS database via a web service or something, and pull that data into a database on the organization's server, and then rebuild the cube. Then that cube would be queried without having to go and connect to the other organization server when viewing the cube. Has anyone else attempted to accomplish something similar? Is HTTP access the standard way to go for this? Or any other possible options? Thanks, and please let me know if you need more info, still unclear on how some of this works.

  Read the article

• exception message getting lost in IIOP between glassfish domains

  - by Michael Borgwardt

  I'm running two glassfish v2 domains containing stateless session EJBs. In a few cases, an EJB in one domain has to call one in the other. My problem is that when the called EJB aborts with an exception, the caller does not receive the message of the exception and instead reports an internal error that is not helpful at all in diagnosing the problem. What happens seems to be this: At the transport layer, a org.omg.CORBA.portable.ApplicationException is created,which already loses all detail information about the exception except its class. Inside com.sun.jts.CosTransactions.TopCoordinator.get_txcontext(), the status of the transaction ass rolled back causes a org.omg.CosTransactions.Unavailable to be thrown, which gets wrapped and passed around a few times and eventually results into this error being displayed to the user: org.omg.CORBA.INVALID_TRANSACTION: vmcid: 0x0 minor code: 0 completed: No at com.sun.jts.CosTransactions.CurrentTransaction.sendingRequest(CurrentTransaction.java:807) at com.sun.jts.CosTransactions.SenderReceiver.sending_request(SenderReceiver.java:139) at com.sun.jts.pi.InterceptorImpl.send_request(InterceptorImpl.java:344) at com.sun.corba.ee.impl.interceptors.InterceptorInvoker.invokeClientInterceptorStartingPoint(InterceptorInvoker.java:271) at com.sun.corba.ee.impl.interceptors.PIHandlerImpl.invokeClientPIStartingPoint(PIHandlerImpl.java:348) at com.sun.corba.ee.impl.protocol.CorbaClientRequestDispatcherImpl.beginRequest(CorbaClientRequestDispatcherImpl.java:284) at com.sun.corba.ee.impl.protocol.CorbaClientDelegateImpl.request(CorbaClientDelegateImpl.java:184) at com.sun.corba.ee.impl.presentation.rmi.StubInvocationHandlerImpl.privateInvoke(StubInvocationHandlerImpl.java:186) at com.sun.corba.ee.impl.presentation.rmi.StubInvocationHandlerImpl.invoke(StubInvocationHandlerImpl.java:152) at com.sun.corba.ee.impl.presentation.rmi.bcel.BCELStubBase.invoke(BCELStubBase.java:225) Is there anything I can do here to preserve information about the actual cause of the problem?

  Read the article

• Test multiple domains using ASP.NET development server

  - by Pete Lunenfeld

  I am developing a single web application that will dynamically change its content depending on which domain name is used to reach the site. Multiple domains will point to the same application. I wish to use the following code (or something close) to detect the domain name and perform the customizations: string theDomainName = Request.Url.Host; switch (theDomainName) { case "www.clientone.com": // do stuff break; case "www.clienttwo.com": // do other stuff break; } I would like to test the functionality of the above using the ASP.NET development server. I created mappings in the local HOSTS file to map www.clientone.com to 127.0.0.1, and www.clienttwo.com to 127.0.0.1. I then browse to the application with the browser using www.clinetone.com (etc). When I try to test this code using the ASP.net development server the URL always says localhost. It does NOT capture the host entered in the browser, only localhost. Is there a way to test the URL detection functionality using the development server? Thanks.

  Read the article

• Manually start session with specific id / transitioning session cookie between domains

  - by deceze

  My host requires me to use a different domain for SSL secured access (shared SSL), so I need to transition the user session between two domains. One part of the page lives at http://example.com, while the SSL'd part is at https://example.hosting.com. As such I can't set a domain-spanning cookie. What I'm trying to do is to transition the session id over and re-set the cookie like this: http://example.com/normal/page, user clicks link to secure area and goes to: http://example.com/secure/page, which causes a redirect to: https://example.hosting.com/secure/page?sess=ikub..., which resurrects the session and sets a new cookie valid for the domain, then redirects to: https://example.hosting.com/secure/page This works up to the point where the session should be resurrected. I'm doing: function beforeFilter() { ... $this->Session->id($_GET['sess']); $this->Session->activate(); ... } As far as I can tell this should start the session with the given ID. It actually generates a new session ID though and this session is empty, the data is not restored. This is on CakePHP 1.2.4. Do I need to do something else, or is there a better way to do what I'm trying to do?

  Read the article

• Grails - Where to store properties related to domains

  - by GalmWing

  This is something I have been struggling about for some time now. The thing is: I have many (20 or so) static arrays of values. I say static because that is how I'm actually storing them, as static arrays inside some domains. For example, if I have a list of known websites, I do: class Website { ... static websites = ["web1", "web2" ...] } But I do this just while developing, because I can easily change the arrays if needed, but what I'm going to do when the application is ready for deployment? In my project it is very probable that, at some point, these arrays of values change. I've been researching on that matter, one can store application properties inside an external .properties file, but it will be impossible to store an array, even futile, because if some array gets an additional value, the application can't recognize it until the name of the new property is added where needed. Another approach is to store this information in the database, but for some reason it seems like a waste to add 20 or more tables that will have just two rows, an id and a name. And the last option, as far as I know, would be an XML, but I'm not very experienced with those. It seems groovy has a way of creating and reading XML files relatively easy, but I don't know how difficult would be to modify an XML whose layout is predefined in the application. Needless to say that storing them in the config.groovy is not an option since any change will require to recompile. I haven't come across some "standard" (maybe a best practice?) way of dealing with these. So the questions is: Where to store these arrays?

  Read the article

• Client-side user custom CSS single file for overriding multiple domains

  - by Cawas

  This is for using in Safari, though it could probably be used on Firefox as well. In Chrome you have to add a plugin anyway (which generally allow for custom CSS per domain), and Opera already allows this to be done without needing any CSS. But while it's for customizing on the client-side, it's also a pure CSS question. So I'm using no plugins here. So, again, I got a custom CSS code (easily) working for all domains. Now I want to get specify CSS code for each domain. All with just 1 CSS file that's being loaded by Safari. Over the web and googling, I've found two ways to supposedly do this, but none actually worked. They're both documented on userstyles.com: @-moz-document domain("your-domain.com") { }. This would be perfect, since I can have several tags like that and just choose which style will be loaded for which domain. It just doesn't work. @namespace is quite confusing and I've tried every variation I could think of. None worked.

  Read the article

• Run SQL Scripts from C# application across domains

  - by Saravanan AR

  To run sql script files from my C# application, I referred this. http://social.msdn.microsoft.com/Forums/en/adodotnetdataproviders/thread/43e8bc3a-1132-453b-b950-09427e970f31 it works perfectly when my machine and DB server are in the same domain. my problem now is, i'm executing this console application from a machine (name: X, user id: abc, domain D1). the SQL server is in another machine (name: Y, domain: D2). I use VPN to connect to machine Y with user id 'abc' and work on SQL in remote desktop. how can i run sql scripts across domains? I use this in a custom activity in my TFS build template as the last step to run the scripts in the target database which is in domain D2. i'm getting this error: "A network-related or instance-specific error occurred while establishing a connection to SQL Server. The server was not found or was not accessible. Verify that the instance name is correct and that SQL Server is configured to allow remote connections. (provider: SQL Network Interfaces, error: 26 - Error Locating Server/Instance Specified)" is it possible to connect to the DB using the admin user ID of machine Y in domain D2?

  Read the article

• Best Practices - Core allocation

  - by jsavit

  This post is one of a series of "best practices" notes for Oracle VM Server for SPARC (also called Logical Domains) Introduction SPARC T-series servers currently have up to 4 CPU sockets, each of which has up to 8 or (on SPARC T3) 16 CPU cores, while each CPU core has 8 threads, for a maximum of 512 dispatchable CPUs. The defining feature of Oracle VM Server for SPARC is that each domain is assigned CPU threads or cores for its exclusive use. This avoids the overhead of software-based time-slicing and emulation (or binary rewriting) of system state-changing privileged instructions used in traditional hypervisors. To create a domain, administrators specify either the number of CPU threads or cores that the domain will own, as well as its memory and I/O resources. When CPU resources are assigned at the individual thread level, the logical domains constraint manager attempts to assign threads from the same cores to a domain, and avoid "split core" situations where the same CPU core is used by multiple domains. Sometimes this is unavoidable, especially when domains are allocated and deallocated CPUs in small increments. Why split cores can matter Split core allocations can silenty reduce performance because multiple domains with different address spaces and memory contents are sharing the core's Level 1 cache (L1$). This is called false cache sharing since even identical memory addresses from different domains must point to different locations in RAM. The effect of this is increased contention for the cache, and higher memory latency for each domain using that core. The degree of performance impact can be widely variable. For applications with very small memory working sets, and with I/O bound or low-CPU utilization workloads, it may not matter at all: all machines wait for work at the same speed. If the domains have substantial workloads, or are critical to performance then this can have an important impact: This blog entry was inspired by a customer issue in which one CPU core was split among 3 domains, one of which was the control and service domain. The reported problem was increased I/O latency in guest domains, but the root cause might be higher latency servicing the I/O requests due to the control domain being slowed down. What to do about it Split core situations are easily avoided. In most cases the logical domain constraint manager will avoid it without any administrative action, but it can be entirely prevented by doing one of the several actions: Assign virtual CPUs in multiples of 8 - the number of threads per core. For example: ldm set-vcpu 8 mydomain or ldm add-vcpu 24 mydomain. Each domain will then be allocated on a core boundary. Use the whole core constraint when assigning CPU resources. This allocates CPUs in increments of entire cores instead of virtual CPU threads. The equivalent of the above commands would be ldm set-core 1 mydomain or ldm add-core 3 mydomain. Older syntax does the same thing by adding the -c flag to the add-vcpu, rm-vcpu and set-vcpu commands, but the new syntax is recommended. When whole core allocation is used an attempt to add cores to a domain fails if there aren't enough completely empty cores to satisfy the request. See https://blogs.oracle.com/sharakan/entry/oracle_vm_server_for_sparc4 for an excellent article on this topic by Eric Sharakan. Don't obsess: - if the workloads have minimal CPU requirements and don't need anywhere near a full CPU core, then don't worry about it. If you have low utilization workloads being consolidated from older machines onto a current T-series, then there's no need to worry about this or to assign an entire core to domains that will never use that much capacity. In any case, make sure the most important domains have their own CPU cores, in particular the control domain and any I/O or service domain, and of course any important guests. Summary Split core CPU allocation to domains can potentially have an impact on performance, but the logical domains manager tends to prevent this situation, and it can be completely and simply avoided by allocating virtual CPUs on core boundaries.

  Read the article

< Previous Page | 19 20 21 22 23 24 25 26 27 28 29 30  | Next Page >