-
-
Need feedback for Ryzen 9950X build
Hello all. I am building my first Unraid server and need your feedback for my build. I will be using the server for running lots of Docker containers (Jellyfin, *arr apps, qbittorrent, Nextcloud, Immich, etc) , running several VMs, running machine learning and AI LLMs (ollama). At first I am planning to start with 1 Parity drive + 1 data drive but in the future I will eventually upgrade to 2 Parity + 10 data drives (The maximum HDDs the Jonsbo case supports) . The mobo has 4 SATA ports so I added a LSI LSI00301 (9207-8i) x8 HBA card to add 8 more ports for a total of 12. 10Gb network is important as well so I added a Intel X550-T2 x4 NIC. Since I heard Intel chips is recommended for Unraid. I won't be using the onboard Marvell LAN port. I'm planning to use all 3 of the PCI-e slots , one occupied by the GPU, the second is occupied by the HBA, and the third is occupied by the NIC, so it will be running at x8/x8/x4. Additionally I'm planning to use the 2 M2 SSDs for cache and docker appdata/VMs. I'm wondering if there will be any PCIe lane or bandwidth limitations or any kinds of bottlenecks with this setup. The mobo I chosen is one of the few motherboards that seems to handle it. ECC is also important for me, so I chose a Kingston 2x32GB memory that's from the mobo QVL. I want a power supply with a high efficiency and can also support 12HDDs so I chose a 1000w 85Plus platinum , but I'm open to a titanium one. I also included a UPS for it. My goal is for this server to be reliable and last me 5 or even 10+ years while also being powerful and future proof. Here is the link to the parts list. Thanks very much! https://newegg.io/93b000e
-
Introducing a Script to Bridge Overseer and Jellyfin for Enhanced Streaming Experience
I love this idea , is it still in development?
-
First unRAID build parts list
Hello all. I am building my first Unraid server and need your feedback for my build. I will be using the server for running lots of Docker containers (Jellyfin, *arr apps, qbittorrent, Nextcloud, Immich, etc) , running several VMs, running machine learning and AI LLMs (ollama). At first I am planning to start with 1 Parity drive + 1 data drive but in the future I will eventually upgrade to 2 Parity + 10 data drives (The maximum HDDs the Jonsbo case supports) . The mobo has 4 SATA ports so I added a LSI LSI00301 (9207-8i) x8 HBA card to add 8 more ports for a total of 12. 10Gb network is important as well so I added a Intel X550-T2 x4 NIC. Since I heard Intel chips is recommended for Unraid. I won't be using the onboard Marvell LAN port. I'm planning to use all 3 of the PCI-e slots , one occupied by the GPU, the second is occupied by the HBA, and the third is occupied by the NIC, so it will be running at x8/x8/x4. Additionally I'm planning to use the 2 M2 SSDs for cache and docker appdata/VMs. I'm wondering if there will be any PCIe lane or bandwidth limitations or any kinds of bottlenecks with this setup. The mobo I chosen is one of the few motherboards that seems to handle it. ECC is also important for me, so I chose a Kingston 2x32GB memory that's from the mobo QVL. I want a power supply with a high efficiency and can also support 12HDDs so I chose a 1000w 85Plus platinum , but I'm open to a titanium one. I also included a UPS for it. My goal is for this server to be reliable and last me 5 or even 10+ years while also being powerful and future proof. Here is the link to the parts list. Thanks very much! https://newegg.io/93b000e
-
Motherboard suggestions
I'll probably get one. What about for a motherboard?
-
garnet672 joined the community
-
Motherboard suggestions
Hello all. I'm building an unraid box after dealing with a dead NAS(ran Jellyfin and media backups). I'm having a difficulty finding a motherboard for my needs. Primary needs off the box will be: mATX or smaller form factor Intel CPU ECC support At least one 10Gbe port Support Nvidia GPU for transcoding 8 SATA ports but possible to expand to 12 in the future M2 slots I'll be using this for docker containers , VMs , media transcoding and possibly running AI models/machine learning.
garnet672
Members
-
Joined
-
Last visited