Unraid Issues New Build


Recommended Posts

Thanks again for your replies.

 

So since I have posted last I made a lot of changes.   I started by contacting Asus for the motherboard to make sure I was using the fastest PCIe ports on the board for the 2 Sata controller cards.  I moved them around to make them all the same.   

 

I took out the Network bond between Eth0 and Eth1 and I created a new bond with Eth1 & 2 because those 2 NICs are in the same dual card I installed.   

 

So now my network is 

 

Eth0 single NIC on Network - 192.168.2.0/24

 

Bond Eth1+Eth2 Single card 2 NICS on Network - 192.168.10.0/24

 

The problem is still there.   I have all Streaming being done over the 192.168.10.0/24 network on a 2gb bond connection.   I use Eth0 for internal use only and it is used to do the downloads for my SAB and newsgroups.   

 

If the Cache drive starts a move which I have set to every hour.  Anything I am watching will pause and start to buffer to a point that you can't watch it.    

 

Also if more than 2 things is streaming at the same time everything buffers.   

 

On my old Unraid server which was much slower than this one I have never had this problem.  I could stream 4 steams at once with no buffering..

 

 

So I can't figure out why any data movement is causing so much overhead.

 

Link to comment

One thing to remember that when you change your server configuration and still have the same problem, it is wise to attach a new Diagnostics file which contains the time frame during which you had the problem.  You have to realize that you seem to have an unique problem.  (If this were a common problem/issue, this board would be have hundreds of threads similar in nature to yours.) 

 

One thing I would be tempted to try is to pull that ethernet card and set the server up using a single ethernet port.  That would give you a setup virtually identical to what a lot of users employ.  If the problem exists then, it is something besides the bonding setup. 

 

By the way, what are you using all of that i7 processing power for?  It seems to be overkill for a simple server setup...

Link to comment

I have actually considered building a 16 core Threadripper for my unraid server.  I am still maxing out my CPU.

I am running a i7 4790k with 4 cores with HyperThreading.

 

In side my house I have a TheaterRoom with 4k UHD Setup.  All my action and Marvell movies I have in full 4K content at about 60gb per mkv.  Most of these I can direct stream with the Nvidia Shield.  However I have all of my family and friends that stream from me as well.  My kids aren't with me half of the time and when they are away they stream from me.   During a busy time I can have 4-6 people outside of my house streaming from my Plex server.   Most of the content is transcoded as most of them don't have great clients and don't support direct play.  Also I use a HomeRun TV box setup in my plex server that does live TV and DVR service.

 

So my Unraid server runs all these Dockers for me.   I have assigned each docker to CPU's so that Plex is on it's own CPU cores.

 

I don't think the problem is with my NIC's anylong.  As long as the data is being written to the Cache drive first everything is smooth.  As soon as the Cache drive starts moving data to the array is when everthing stops.

 

Hope that makes sense.   I have attached a new diag file.

medianas-diagnostics-20190326-1425.zip

  • Like 1
Link to comment

I looked through your Diagnostics file and saw that in the syslog, there were a lot of errors that involved disk 1.  This is the beginning of that sequence: 

Mar 23 01:25:28 MediaNAS kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
Mar 23 01:25:28 MediaNAS kernel: ata3.00: irq_stat 0x40000001
Mar 23 01:25:28 MediaNAS kernel: ata3.00: failed command: READ DMA EXT
Mar 23 01:25:28 MediaNAS kernel: ata3.00: cmd 25/00:40:a8:f2:98/00:05:81:01:00/e0 tag 23 dma 688128 in
Mar 23 01:25:28 MediaNAS kernel:         res 51/40:00:08:f6:98/00:00:81:01:00/00 Emask 0x9 (media error)
Mar 23 01:25:28 MediaNAS kernel: ata3.00: status: { DRDY ERR }
Mar 23 01:25:28 MediaNAS kernel: ata3.00: error: { UNC }
Mar 23 01:25:28 MediaNAS kernel: ata3.00: configured for UDMA/133
Mar 23 01:25:28 MediaNAS kernel: sd 3:0:0:0: [sdd] tag#23 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
Mar 23 01:25:28 MediaNAS kernel: sd 3:0:0:0: [sdd] tag#23 Sense Key : 0x3 [current] 
Mar 23 01:25:28 MediaNAS kernel: sd 3:0:0:0: [sdd] tag#23 ASC=0x11 ASCQ=0x4 
Mar 23 01:25:28 MediaNAS kernel: sd 3:0:0:0: [sdd] tag#23 CDB: opcode=0x88 88 00 00 00 00 01 81 98 f2 a8 00 00 05 40 00 00
Mar 23 01:25:28 MediaNAS kernel: print_req_error: I/O error, dev sdd, sector 6469251752
Mar 23 01:25:28 MediaNAS kernel: md: disk1 read error, sector=6469251688
Mar 23 01:25:28 MediaNAS kernel: md: disk1 read error, sector=6469251696
Mar 23 01:25:28 MediaNAS kernel: md: disk1 read error, sector=6469251704

I am not sure what is going on here (as I no Guru when it comes to these types of problems) but you could could google things like   failed command: READ DMA EXT   and    print_req_error: I/O error, dev sdd, sector 6469251752    Or wait until someone more knowledgeable comes along.  

Link to comment
  • 1 year later...

I know this is an old thread. However, thought it would be usefull for for others coming across this.

When you have working with multiple networks on a server (any computer), you have to think about routing. There is no guarantee that the return traffic is going back on the same interface as the request arrives on. Traffic may arrive on nic1 but return traffic may go back via nic2. On linux, you can control this in detail by createing unique routing tables per nic etc. You will however get a very complicated network setup.... if you have multiple networks, you can use specific routing where network a, b, and c are using nic1 (with routing entries in unraid) and network d, e, and f are using nic2 (with rouging for that in unraid). 

 

However, I would strongly suggest, using LACP and bond your network cards to that (must be same speed).

(sure, you may have to buy a managed switch for a couple o hundred dollars, but worth every penny)

If need to, configure QOS on your switch.

Sending data to Unraid, while streaming, should not be a problem as you are looking at dataflows in two different directions.

Make sure you have a cache drive that is minimum twise the speed of your network, and configure yours shares to use the cache drive to write new content. (I.e. you stream from your hdd, write new content to cache)

 

Rule nr 1. Keep it simple!

Rule nr 2. Simplify, i possible.

Rule nr 3. Go to 1.

 

/Martin

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.