Jump to content

Hoopster

Members
  • Content Count

    2027
  • Joined

  • Last visited

  • Days Won

    10

Hoopster last won the day on October 9

Hoopster had the most liked content!

Community Reputation

270 Very Good

2 Followers

About Hoopster

  • Rank
    Advanced Member

Converted

  • Gender
    Male
  • Location
    Utah USA

Recent Profile Visitors

2276 profile views
  1. As the owner of a Node 304, Lian-Li PC-Q25B and Silverstone CS380 all of which have housed unRAID servers, I can tell you that each have their plusses and minuses as is always the case. Node 304: - a great case if you do not swap disk drives often and do not need easy access to the internals of the system. It's a bit of a pain to work with when you need to make changes - very compact case (but you already know all about it) - currently houses my backup unRAID server which just sits in a cool corner of the basement and gets powered on/off by IPMI when needed Lian-Li PC-Q25B - housed my original unRAID server for many years (it is currently empty awaiting the need for another small server build) - drives are very accessible and, although it has "only" five hot-swap bays, it can house eight or more drives if you double stack on the bottom tray. I read an article once in which the author demonstrated how he had put 11 drives in the case (9 3.5" and 2 2.5" SSDs - cooling is adequate but is limited to one 140 mm fan and one 120mm fan - case internals are easily accessible as both sides pop off - power supply covers CPU and limits cooler height to 80mm (I used a 65mm Noctua "downdraft" cooler in that case) Silverstone CS380 (note this is NOT the DS380 nor the CS381) - has housed my main server for about a year - fairly compact for a case that can handle Mini-ITX, mATx and ATX motherboards - cooling issues are easily overcome with case mods documented by dmacias in these forums - can take up to 8 3.5 HDDs in front-facing hot-swap bays. I love the easy accessibility to the drives - has two 5.25" bays above the 3.5" hot swaps. I have a 4x2.5" SSD hot swap in one and an optical drive in the other for MakeMKV. These could be modified for additional 3.5 HDDs, if desired - easy accessibility to MB and other internals - the CPU/hard drives (after the mods) are cooler and the case is quieter than my build was in the Lian-Li (my saga in getting there is documented in the above linked thread) - price has gone up somewhat recently (I got mine for $139 last year) I can't tell you about the other options you listed, but, I really like the CS380 after the mods and the Lian-Li PC-Q25B is a great Mini-ITX case if you can find one. Unfortunately, I sold my PC-Q25B spare last year. Many users are looking for towers with all 5.25" bays so they can put 3 or more 5-bay hot-swap cages in them, but, those are incredibly hard to find as well. Good luck. Hope you find something that works for you.
  2. Depends on how you use the cache drive and how large it is. I have only a 250GB cache drive but I do not cache any writes to array shares with it. It is only houses the appdata (docker container share) and is the initial landing spot for downloads. I have another 500GB SSD as an unassigned device which I use for hosting VMs and Plex transcoding. You could theoretically do all of the above plus write caching on a single large SSD but it may be busy. 😁
  3. Just because you are streaming 1080p content to a 4K TV does not mean the Plex is going to upscale to 4K resolution. Your TV (or certain client devices such as the nVidia Shield) may attempt to upscale, if it does not, Plex will always try to direct play the files in their native format/resolution if the streaming/playback device supports it. On the client side, you can tell Plex to transcode to a particular resolution other than the native resolution of the file, but, any 4K transcoding is going to require about 17,000 passmarks according to Plex, Your CPU has a passmark rating of 18426, so you should be able to get one 4K transcoded stream with barely enough overhead for unRAID itself (generally about 2000 for unRAiD is the recommendation). FYI- generally transcoding goes the other way, for example 4K ---> 1080P or 1080P --> 720P, etc. Basically, the rule to thumb is that Plex will first try to direct stream to the client in the native format, if that format is not supported on the client side, it will transcode. You can always force transcoding and remote streaming usually does involve transcoding. For this reason, may of us maintain separate 4K and 1080p content libraries. 4K content for local streaming to devices that support it natively so it can be direct played and 1080p libraries for remote streaming and streaming to devices without 4K content support. 4K transcoding is very resource expensive for the CPU or GPU depending on your configurataion You should run Plex as a docker container. Dockers are another argument for an SSD cache drive as you should store the appdata share and docker.img on an SSD when possible for better performance and few array disk spinups.
  4. unRAID is hardware agnostic. The only hardware that matters for licensing purposes is the flash drive used to boot unRAID. Change hardware all you want. You can also extend the trial twice for 15 day each.
  5. Yes. In fact, the script I use backs up only new/modified files to the backup server. It DOES NOT delete anything from the backup server that has been deleted since the last backup of the source server. I have another script that I run manually every three or four months to cleanup deleted source files from the backup. That usually gives me time to see if anything has been accidentally deleted which has happened twice. I once lost a whole year of photos on the source server that I was able to recover from the backup server because of the way this is being managed.
  6. 1. Pihole and/or docker issues means your server has no DNS 2. There are no ads to block when running unRAID though Pihole so it's an unnecessary layer of processing (granted, it's likely to be unnoticeable) 3. If router, server and/or Pihole have a misconfiguration, it could result in DNS loops that don't get resolved 4. You have the problem you mentioned with a custom IP address when Pihole is running as a docker on macvlan Some are, in fact, running their unRAID server through Pihole without issue, but, it is really not necessary. I have Pihole as a docker container for backup purposes, but, I am currently running Pihole on a Raspberry Pi. Either way, the configuration is the same. Pihole IP address is the specified DNS in my router so all connected clients go through it, but, unRAID bypasses Pihole and goes direct to Cloudflare for DNS.
  7. It's not a good idea to have Pihole as your DNS for the unRAID server. In Settings --> Network Settings set the DNS for the unRAID interface to some external DNS such as Google, Cloudflare (these are my DNS settings), etc. or your router IP address (depending on how you have it configured).
  8. Edit syslinux.cfg in the syslinux folder on your boot flash drive. Add iommu=pt to the append line to the desired boot mode(s) so it looks something like this: default menu.c32 menu title Lime Technology, Inc. prompt 0 timeout 50 label unRAID OS menu default kernel /bzimage append initrd=/bzroot iommu=pt label unRAID OS GUI Mode kernel /bzimage append initrd=/bzroot,/bzroot-gui label unRAID OS Safe Mode (no plugins, no GUI) kernel /bzimage append initrd=/bzroot unraidsafemode label unRAID OS GUI Safe Mode (no plugins) kernel /bzimage append initrd=/bzroot,/bzroot-gui unraidsafemode label Memtest86+ kernel /memtest
  9. I use the Unassigned Devices plugin and have modified the sample photos backup script in the first post of the plugin thread so that it backs up three shares. The script automatically does the backup when the external USB drive is plugged in.
  10. All my shucked drives are from Easystore enclosures. Early on they were EFZX reds. For about two years they have all had EMAZ white drives in them. The EMAZ drives are definitely Helium filled (at least all the ones I have) as Helium level displays as a SMART attribute on all of them. AFAIK (I do not know from personal experience), I had heard/read that the Elements enclosures also contain EMAZ drives. I have no idea what may be in the MyBook enclosures. I try to stick to Easystore when feasible as I know what I will get with them.
  11. 1 - With the drives you have, the 12TB drive MUST be the parity drive as parity must be as large or larger than the largest data drive. No data drive can be larger than the parity drive(s). Your 4TB, 3TB and 2TB drives can be added to the array as data drives and the 12TB parity drive will allow you to add new drives or replace existing drives with any drive up to 12TB in size. 2 - Depends on how you intend to use the cache drive. If you want to use it for caching writes to the array and you commonly transfer large amounts of data, the 1TB M.2 SSD would likely be the best cache drive option. If you are not caching writes for any shares (the Mover later moves them to the array) and only want to use the cache drive for hosting shares such as appdata (docker containers), system, etc., the 250GB SSD is plenty for this. The 1TB could then be added as an unassigned device and used for hosting VMs. Of course, how you use them is up to you. I have a 250GB SSD as my "cache" drive with no caching; all writes go directly to the array. This cache drive is also the initial landing spot for downloads. I also have a 500GB SSD (soon to be 1TB) as an unassigned device on which I host VMs. I also use it as the landing place for MakeMKV and HandBrake conversions before I move the files to the array. 3. Yes, two VMs can have the same video card assigned to them, but, of course, they cannot use it simultaneously. The first VM you start will "own" the video card until you stop the VM, then, the other VM can use it. If you want to run the VMs simultaneously, you'll need two video cards. 4. Cloud backup, backup to another server, external USB drive (automatic backup via Unassigned Devices plugin and User Scripts). I do all three.
  12. 👍 Glad to hear all is working for you with the correct cable.
  13. Hard to tell from just a picture. You need forward breakout cables. Forward and reverse breakout look the same. On my H310 I had to tape pins 5-6 for drives to be recognized. Since you have tried that perhaps the cables are the issue. Can you find a description of them to verify that they are forward breakout? Sent from my iPhone using Tapatalk
  14. In my case, I have a mix. Some are on br0.3, some are on bridge and some are on host. With my hardware, I only got call traces on br0 so I have no containers on that network. Sent from my iPhone using Tapatalk
  15. I am afraid I do not have an answer for you. I came across the concept of secrets when I was looking into some other facet of Docker a few months ago so I knew what you were talking about, but, I have seen no discussion of it relative to the Docker implementation in unRAID. There is this note in the document I linked: "Note: Docker secrets are only available to swarm services, not to standalone containers. To use this feature, consider adapting your container to run as a service. Stateful containers can typically run with a scale of 1 without changing the container code."