Just generally unsatisfied... maybe i am doing something wrong?

[titanarchon]

Hey everyone. I just built an UNRAID server and I am coming from a Synology. Specs are as follows:

Ryzen 2600

16GB 3200mhz RAM

LSI Logic SAS 9207-8i Storage Controller LSI00301

8 x 8TB Iron Wolfe

1 x 500GB Samsung EVO 860 SSD

Here is the story, i built the machine and installed UNRAID with no issues. Works amazing with no workloads. I have changed nothing from the factory settings and all discs are included with all shares and cached. I have 1Gbps up and down fiber for internet btw. I have Sonarr, Radarr, Sab, and Plex running. I let Radarr do its thing and watch this powerful box drag to a crawl. right now, i have 76 movies in queue and its downloading at 4.6mb, it has 3 completed but not moved to the completed folder, and its just inching along. The interfaces seem to move about ok and Plex doesnt seem to be affected at least with 3-5 transcodes going, but i need to figure this out. My Syno smoked this thing and with 1/8th the hardware. Thank you and other than this i do love the product. I will have to buy pro when i get it but still not convinced on it as of yet. Any help is appreciated.

[Squid]

diagnostics as a start would be needed for anyone to begin to offer suggestions

[titanarchon]

48 minutes ago, Squid said:

diagnostics as a start would be needed for anyone to begin to offer suggestions

Considering it shows I'm a newbie and its my first post, let me know which of the 10k logs you need for diagnostics and ill provide everything you need sir. Thank you for the quick reply!

[Squid]

Just now, titanarchon said:

let me know which of the 10k logs you need for diagnostics

The entire zip file generated by tools - diagnostics

[titanarchon]

1 minute ago, Squid said:

The entire zip file generated by tools - diagnostics

Doing this now. Thank you!

[titanarchon]

OK attached sir. Thank you.

plex-diagnostics-20190126-1105.zip

 

Edited January 26, 2019 by titanarchon

[trurl]

Something wrong with your docker configurations. You have allocated 120G and already used 32G. If you are doing it right you shouldn't need any more than 20G allocated.

 

Have you read the Docker FAQ? Pay particular attention to the section on "Docker Image Filling Up"

 

https://forums.unraid.net/topic/57181-real-docker-faq/

 

You also have some sort of a hardware problem spamming your syslog.

[titanarchon]

This was due to my extremely large Plex library. Around 16tb. THe metadata from Plex took too much space. I had to expand to 120GB to accomodate metadata which i got from that FAQ. Can you elaborate on the hardware issue at all?

[titanarchon]

I mean i could take that down to 60 until it get filled up, but i don't feel that would affect speed.

[Squid]

Best guess is that you're IO bound, quite possibly because what I'm assuming is your Downloads share for sabnzbd (D-------s)

# Share exists on cache,disk1,3,4,7

Which entails a huge performance penalty

(and also noticed via this line)

Jan 26 05:11:00 Plex shfs: share cache full

 

Additionally, this isn't helping things

Jan 26 09:24:39 Plex kernel: pcieport 0000:00:03.1: AER: Multiple Corrected error received: 0000:00:00.0
Jan 26 09:24:39 Plex kernel: pcieport 0000:00:03.1: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)
Jan 26 09:24:39 Plex kernel: pcieport 0000:00:03.1:   device [1022:1453] error status/mask=00001100/00006000
Jan 26 09:24:39 Plex kernel: pcieport 0000:00:03.1:    [ 8] RELAY_NUM Rollover    
Jan 26 09:24:39 Plex kernel: pcieport 0000:00:03.1:    [12] Replay Timer Timeout  

Try looking here https://forums.unraid.net/topic/77406-new-to-unraid-varlog-is-100-full/?tab=comments#comment-715343

2 minutes ago, titanarchon said:

Around 16tb. THe metadata from Plex took too much space.

Don't use plex, but my understanding is that all of the metadata should be stored in /config.  Maybe post your docker run command https://forums.unraid.net/topic/57181-real-docker-faq/#comment-564345

 

[CHBMB]

3 minutes ago, titanarchon said:

This was due to my extremely large Plex library. Around 16tb. THe metadata from Plex took too much space. I had to expand to 120GB to accomodate metadata which i got from that FAQ. Can you elaborate on the hardware issue at all?

There's no way a 16TB plex library needs that much storage for it's metadata.....

 

People on here have much bigger Plex libraries and it shouldn't need 120GB docker image to accomodate it, especially as the metadata is stored in your appdata folder not the docker.img file.

 

I think you have some issues with how you've set up your docker containers.

[trurl]

4 minutes ago, titanarchon said:

This was due to my extremely large Plex library. Around 16tb. THe metadata from Plex took too much space. I had to expand to 120GB to accomodate metadata which i got from that FAQ. Can you elaborate on the hardware issue at all?

Your plex metadata isn't in the docker image. It is in appdata which should be on cache unless you have done something wrong. Lots of people running plex and nobody ever needed to have a docker image larger than 20G.

1 minute ago, titanarchon said:

I mean i could take that down to 60 until it get filled up, but i don't feel that would affect speed.

It should never fill up. It should hold steady at less than 20G. If it increases you are doing it wrong.

 

Post your docker run command as explained in the very first link in that Docker FAQ.

[titanarchon]

OK. You guys are obviously extremely smart and kinda over my head on this. Other than fixing the PCIE issue (i think that card is being limited to x4 instead of its native x8 which ill move it to the other slot in a few and see how that works) what do you all need in order to help me correct what i have done "wrong?" (i dont appreciate that term considering all i have done is pointed to my plex library but say what you want) Thanks for the support guys. 

[CHBMB]

3 minutes ago, titanarchon said:

OK. You guys are obviously extremely smart and kinda over my head on this. Other than fixing the PCIE issue (i think that card is being limited to x4 instead of its native x8 which ill move it to the other slot in a few and see how that works) what do you all need in order to help me correct what i have done "wrong?" (i dont appreciate that term considering all i have done is pointed to my plex library but say what you want) Thanks for the support guys. 

You can call it whatever you want, but I agree something isn't right,  I think you should start with posting your plex docker run command.  Link on how to do so is in my signature. (You may need to enable signatures in the forum)  Alternatively, read the Docker FAQ post in the Docker forum, which is where you can learn how to do things the right way.  It's probably an issue with volume mappings.

[titanarchon]

2 minutes ago, CHBMB said:

You can call it whatever you want, but I agree something isn't right,  I think you should start with posting your plex docker run command.  Link on how to do so is in my signature. (You may need to enable signatures in the forum)  Alternatively, read the Docker FAQ post in the Docker forum, which is where you can learn how to do things the right way.  It's probably an issue with volume mappings.

Ill give that a shot. Thanks for the help. I think its a mapping with Sab seeing as its log is 31gb, not Plex's. Thanks for the help again guys. 

[titanarchon]

Guys i know from the looks of your previous posts this is going to incite some hate, but you all need to work on how you respond to people. Seems like there is just a general disdain towards people asking questions in a lot your posts. We are all here for a common software need and joy of doing these things. Maybe keep that in mind. I am in no way challenging your smarts but rather relying on them for a smooth deployment. I appreciate all your help and think i found the issue in Sab thanks to you guys as well as the PCIE issue. Sometimes researching for ourselves (noobs) leads to more questions than answers. I thank you all again for your time, intelligence, and prompt responses. You may mark this as solved if you would like. 

[CHBMB]

1 minute ago, titanarchon said:

Guys i know from the looks of your previous posts this is going to incite some hate, but you all need to work on how you respond to people. Seems like there is just a general disdain towards people asking questions in a lot your posts. We are all here for a common software need and joy of doing these things. Maybe keep that in mind. I am in no way challenging your smarts but rather relying on them for a smooth deployment. I appreciate all your help and think i found the issue in Sab thanks to you guys as well as the PCIE issue. Sometimes researching for ourselves (noobs) leads to more questions than answers. I thank you all again for your time, intelligence, and prompt responses. You may mark this as solved if you would like. 

I've had a look over the thread and I got to say, nobody has been rude or dismissive, I think you're being overly sensitive.  I am the first to admit I can be rude and dismissive to people when they're rude or ungrateful, but I don't think I'm guilty of that here.   I don't really know what you've taken issue with, but it would be very wise to remember that all three of us that have replied her are volunteers and not employees.  I think you need to work on not being so sensitive.

 

If you've read back through my 9700+ posts I suspect you'd find the vast majority of them are helpful, so whilst I have heard your opinion I am going to disagree with it.

 

And as for @Squid and @trurl I would say the above could equally applied to them. 

[limetech]

2 hours ago, titanarchon said:

you all need to work on how you respond to people.

It's very difficult to ascertain the tone of forum posts.  I'm sure you've noticed most forums and sites tend to devolve very quickly into nastiness, and sometimes people think that's happening here.  We work very very hard at making this a friendly place, especially for new-comers.  All our mods and key contributors are very patient and helpful, even when asked the same thing over and over.  I'm 99% sure no rudeness was intended.

[Squid]

3 hours ago, titanarchon said:

I think its a mapping with Sab seeing as its log is 31gb

https://forums.unraid.net/topic/57181-real-docker-faq/?tab=comments#comment-564326

 

[jebusfreek666]

@Squid, @trurl, @CHBMB   

 

wow.....

 

On behalf of the less vocal in this community who rarely have anything to give you guys but our problems, I just wanted to say thanks. Sincerely, thank you guys for the time you give in helping others. Even when you get BS in return for your generosity, you continue to contribute massively to the community. You are appreciated! 

[Squid]

4 hours ago, jebusfreek666 said:

Even when you get BS in return

Everybody gets in bad moods (myself included).  Sometimes that does show on our posts, and any spanking that we get from users are sometimes deserved.  But we're still here, still volunteering our time out of the enjoyment that we get supporting our favourite project...

[electron286]

If the problems have been resolved GREAT!  but...  I thought from the initial post that Radarr was causing an issue, as that was when the crawl popped up...  In looking into it, and some related Radarr posts, it seems there have been some memory leaks, that should be fixed now in the current builds, and also some configuration problems some people have seen performance issues with, the biggest one being;

 

  Quality profiles - Be sure to set your CUTOFF settings appropriately for EACH Quality profile to allow Radarr to stop looking for better versions of downloads.

 

  Followed by not setting up Delay Profiles, to allow for a longer delay before looking again for matching downloads.

 

  I know some people really do not like the word WRONG, so that may be the source of the issue as seeing the replies as unfriendly.  While I also use the word, I try to use the words incorrect, or incorrectly, when writing to an unknown audience.  Wrong is often taken personally, regardless of how it was used or intended.  Add to that how many times the word wrong did appear in multiple postings, and... well you get the general idea.  No matter why the emotions are ignited, it is hard to look at any other part of a conversation without them getting in the way.

 

  The Unraid forums have been the BEST forums I have ever seen as far as keeping things civil, and while we all make mistakes at times, all the guys here really do care and try to help each other.  We all have a love for computers, and the wonderful solution that Unraid has provided for many years, and now the many MORE solutions it provides!  I would like to say WELCOME titanarchon, and I hope you can enjoy your Unraid server and feel better about this forum.

 

  We are all new at some point, and even newbies on some aspects of Unraid with all the changes that have been rolled out over the years.  I am still mostly way back on version 4.5.3 for an example.  I have set up newer versions, but mostly 4.5.3 is my work horse. (NO I WOULD NOT RECOMMEND ANYONE USE 4.5.3 UNLESS THERE IS A REAL NEED, DUE TO THE MANY POTENTIAL USER ISSUES THAT COULD LEAD TO DATA LOSS!  These problems were all resolved nicely in newer releases.)  I am also the ONLY one that touches my old units.  If I had anyone else touch my servers, I would be running newer versions just for the safety and peace of mind.  Anyway, with each new Unraid server I set up for other people, I always use the newest stable version, and always need to learn what changes were made since the last server I set up.  Sometimes the new features do have a bit of a learning curve, (such as setting up shares to work properly in a PLEX Docker..).

[trurl]

7 minutes ago, electron286 said:

I know some people really do not like the word WRONG, so that may be the source of the issue as seeing the replies as unfriendly.  While I also use the word, I try to use the words incorrect, or incorrectly, when writing to an unknown audience.  Wrong is often taken personally, regardless of how it was used or intended.  Add to that how many times the word wrong did appear in multiple postings, and... well you get the general idea.  No matter why the emotions are ignited, it is hard to look at any other part of a conversation without them getting in the way.

Well I am certainly guilty of using that word here. But the OP used it first and I was mostly going with it for that reason. If there was any snark it was all in fun.

 

We really are trying to help.

[electron286]

Performance...

 

  I just saw this thread that also may have also been related to the initial problems...

 

https://forums.unraid.net/topic/76834-please-help-cpu-spiked-to-100-plex-unusable/