Search the Community
Showing results for tags 'GettingStarted'.
Hello and Welcome First, let me make the most important point, Emby Server is fully supported on XPEnology systems running DSM 5.2-5967 or later. If you're running an earlier version, check out the XPEnology forums for advice on how to update. We've had many XPEnology users update without issue and there are members of the community here who will also help, just ask. To reiterate what is detailed under Step #1 on the Synology setup page (https://emby.media/nas-server.html), please make sure you use this URL as your package source in Package Center:- https://synology.emby.media/?package_repository=360efc6e-de72-4073-b603-2bfbd7001586 Emby Server Won't Run Now, if you've installed the package successfully but it won't run, don't panic! We almost certainly know why that is, and I'll explain in detail below. If you're impatient and just want to get going, please run the two utilities listed below... Run Diagnostics : https://github.com/MediaBrowser/Wiki/wiki/Synology-:-How-to-Run-Diagnostics Send Support Logs : https://github.com/MediaBrowser/Wiki/wiki/Synology-:-How-to-Send-us-Support-Logs ...once the send logs utility has completed, post your details here along with the send logs file reference. What's the Problem? The CPU type and architecture reported by DSM when running on XPEnology is fake, which is due to it 'pretending' to be a DS3615xs. With the most recent phase of package development, we've optimised the toolchains we use to build the packages on an architecture-by-architecture basis, so as to get the best performance out of the hardware. This causes problems on the XPEnology front, as it can theoretically be running on any Intel or AMD x86-64 processor available. We have no way of knowing that it's an XPEnology system or the underlying processor type. The hardware that Synology groups under the 'bromolow' architecture, which includes the DS3615xs, contain either an Intel Xeon E3-1230 v2 or Intel Core i3-4130 processors. These are based on the 'ivybridge' and 'haswell' microarchitectures repectively. In terms of general CPU capability, this maps out as follows:- ‘ivybridge’ MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES, PCLMUL, FSGSBASE, RDRND, F16C ‘haswell’ MOVBE, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2, F16C So, for Synology hardware identifying as 'bromolow', we target the 'ivybridge' microarchitecture as the lowest common denominator. If the processor you're running XPEnology on supports these CPU features, then you're going to 'get away with it' and it will work. We may however, not be taking advantage of all the capabilities your processor has to offer. The first time we came across an issue was with the Intel Pentium G3258 processors, as these do not support the AVX feature (see above). This is the point where we decided to split the XPEnology and Synology builds and give XPEnology a dedicated repository, which you'll see in the setup guide on the main website. Our default target now for XPEnology systems is the 'nehalem' microarchitecture, which supports the following features:- ‘nehalem’ MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT Later on, we came across the Intel Core2 Duo E7200, which doesn't support SSE4.2 or POPCNT, again causing Emby Server to not start. The lowest common denominator of Core2 based processor doesn't support SSE4.1 either, so we could have dropped our baseline from 'nehalem' to 'core2', which would have given:- ‘core2’ MMX, SSE, SSE2, SSE3, SSSE3 However, our objective is to make Emby Server run as well as it possibly can, so this wasn't really a path we were willing to take; although it would have certainly been the easier one. For that reason, we enhanced our package server to allow a package architecture to be specified as part of the package source and added the ability to create additional platforms to our toolchain, in this example for Core2 processors with SSE4.1. This gives us the flexibility to support processors that are less capable than our default ('nehalem') and further optmise for processors that are better than our default in the future. Best - James