I thought I would share my experience with you fellow QNAP owners who use Plex. I am running Plex as a native app on the QNAP (TVS-1282T2). I am running WD Reds (6TB) drives as the main store drives in the unit. I am also running an expansion unit (TX800P) on the NAS using an assortment of WD Reds and some other random drives (4TB).
The Plex Media Server was installed on the main array in the normal system area. The Database was getting larger as more movies and TV shows were added (1500+ movies, 4000+ Episodes of TV shows) and had noticed that at times the system was a little slow to respond. It seemed that as the DB got larger the read times for the Database were slowing down. I had also read this elsewhere on this forum about larger databases and read times.
Talking to @ChuckPa he put me on to moving the PMS App and database to SSD drives to speed up access. Previously you could do this with PMS but was a bit of a tightrope walking act to move the files and keep continuity of operation. With the latest builds of PMS you can do this pretty easily within the QNAP operating system.
I now have two M.2. 512Gb SATA SSD drives installed internally on the QNAP configured in RAID 1 as a static volume (no need for Drive Pool as these drives are dedicated to Plex. Static volume gives you higher write and read speeds) and I migrated the PMS to this new disk config in the app centre (to do this you right click on the app in app centre and then select migrate. It will allow you to pick the new location as long as you have the SSD’s set up as a volume)
This has made PMS so much more responsive and fast , even to the external WAN clients that access Plex. Transcoding is quicker, menu response is much quicker, it’s a bit of a different beast now. If you are dealing with larger databases then this might be an option for you.
Thanks to @ChuckPa for the heads up and advice. The next project will be to move the QNAP OS to an SSD for a faster response, although that will be a bit more difficult. To get some real good speeds instead of using the M.2. slots on board the QNAP motherboard which are only SATA, I will look at using the PCIe slots in the unit and using the faster PCIe SSD’s mounted on a QNAP expansion card.
made a backup (screenshots) of all my configurations and a QNAP backup of the config.
With the system having been installed on the HD array, this would be easy.
A. Power off and pull all the drives
B. Install the M.2 cards and put it all back together - EXCEPT the drives
C. Using the QNAP utility, Download and install QTS on the SSD - STATIC volume.
D. Configure the SSD as RAID 1 (mirror)
E. Verify all networking is in place as it was. Correct as you need to.
F. Everything ready, power off
G. Insert the main HDD array only
H. Power up
I. Observe QTS now finds this as CACHEDEV2_DATA. Remember this fact.
You’re now up with a SSD OS, HDD data configuration.
The pain is just beginning.
Recreate all your shares by pointing (manual) to the existing directory on CACHEDEV2_DATA
Configure all your network sharing (SMB, NFS, etc)
Re export all the shares you previously had exported.
Install the other apps you had.
In Storage management, make sure SSD trim is enabled and set to the schedule of your liking. I use Weekly trim.
All of the above is why you have screenshots as backup
In retrospect, had I known it would be this much better after installing the SSDs, I would have done it upfront.
As RAID-1 we get 2x read speed (they self-stripe). Both are written in parallel but read stripes.
For the final step, if you have more drives to install, Insert them in the sequence you wish defined as. These will now be CACHEDEV3_DATA, CACHEDEV4_DATA, etc.
If you have a 2.5" SSD for read-cache, insert and enable it.
The QNAP OS itself resides on the internal flash device. It’s only about 200 MB in ZIP format so I would guess 300-400 MB when installed.
By installing the OS to the flash, no HDD space is used other than some swap space (64 GB swap partition on my system) and the QPKG directory (for packages we install). Temp is a RAMdisk.
I plan out my storage and set max volume size based on the biggest drives and count I intend to put in the box.
For SSD, I keep it simple. The SSD is a single-disk volume so there’s no reason to make it complex. It’s size won’t grow; I’m not going to snapshot the SSD. Go from there.
If you want volume snapshots, go with Thick. If not, and won’t be changing config often. go with static.
When I created my corrent main array ( 8x WD Red Pro ), I set the inode size to support a 127TB volume. The biggest I will put in the box are 14TB drives. By then, I’ll have a new NAS.
As I said initially, my CACHEDEV1_DATA is RAID 1. If I’m doing RAID 1, it makes sense to make it static. I make a regular backup of it when I need to. I’m old school.. simple, clean backup
