October 6, 201411 yr Hi, Over the weekend I decided to get rid of unraid v5 and start using V6 and docker for my apps (plex, nzbget, nzbdrone). The upgrade went well to V6 then I reformatted my cache drive (where I installed all my apps as cache only) to btrfs in anticipation of Docker apps. I then reinstalled all my apps (to mnt/user/appdata - as a cache only share) using the docker templates provided in 6.0-beta10a. This again went fairly well, although I did notice that I had to run Plex refresh a couple of times to pick up all my movies and Tv shows and it was very slow. My problem is as follows - I have noticed a considerable drop off in speed/performace of Plex. What I mean by this is when I used 5.5 and plugins and I browsed to main menu/Movies/Tv sections in Plex it would switch to them near instantaneously but now with the docker apps it can take up to 10-15 seconds to open up. It might not sound like much but when you are used to it being very quick this does get annoying. Could anything I have done caused this? Cheers
October 6, 201411 yr Were you running cachedirs on 5.x? Are you running cachedirs on the 6.x series? Be sure to reinstall cachedirs if you have not done so already.
October 6, 201411 yr Author Thanks for the reply but no I wasn't running anything called cachedirs on 5.x and am not running it now on 6.x.
October 6, 201411 yr I'm running Dockers too for all my Apps need and it's blazing fast here... no apparent slowdown. In fact, it's fastttttt !
October 7, 201411 yr I had this same issue when i first moved to v6, however after a few days plex returned to normal speeds and has been fine ever since. Could be to-do with regenerating all the media info, and parity checks etc. Give it a few days, and you might find the problem solves itself.
October 7, 201411 yr Author I switched out my cache drive yesterday and replaced it with a 256gb SSD I had lying around. It took a while to rebuild my plex library but now that it has it is back to being fast. Actually it's probably faster now.
October 7, 201411 yr When you create a new database/add a lot of data to Plex, make sure you kick off an "Optimize". If it cleared up over time, then it might be because Plex does this automatically every week. I got a huge slowdown after my initial move until I did this manually.
October 9, 201411 yr My experience, also, is that there may be some performance issues with docker, or latest betas. Sorry for the historical barration, but I think that it provides evidence with respect to the current topic. Please note that I am not running Plex. I am using the apcupsd plugin and a second machine, running apcupsd, is slaved off of it, using a network connection to the unRAID instance of the daemon. With beta 4, I started using a VM to offload the functions of some plugins - namely: LogitechMediaServer, Deluge, Minidlna, CouchPotato and MySQL (MariaDB). These were running on one instancr of IB's ArchVM. I also set up a second instance of ArchVM, with a desktop, in order to run jdownloader. With the upgrade from 5.04 (I think, or was it 5.05?), to beta 4, I abandoned use of cache_dirs. I noticed that the apcupsd slave machine was, very occasionally, loosing connection to the master apcupsd on the unRAID box, for a few seconds at a time. Then there was some issue with Xen VMs starving Dom0 of resource, for which the fix was to pin a cpu to Dom0. This I did, leaving the other 7 cpus for the VMs. This worked well, and the apcupsd 'lost connection' issue vanished. Then Joe provided a cache_dirs compatible with V6, which I installed. I immediately noticed that the slave apcupsd machine started to loose connection with the master, with the connection being lost for several minutes at a time (20 - 30 minutes). I pinned a second cpu to Dom0, and all was well once more - no more apcupsd lost connections (now running beta 5). I avoided beta 6 bexause of teh problem with DomU networking. I then upgraded -> beta 7 -> beta 8 and started playing with docker. Once I had a couple of dockers running, I noticed that the apcupsd 'lost connection' issue came back. I didn't worry too much. believing that it would resolve once I had all my ArchVM daemons running as dockers and could shut down the ArchVM. I have now reached that position, have no VMs running (although I'm still booting with Xen support). The apcupsd 'lost connection' error persists, even though unRAID (Dom0) now has all eight cpus available to it. So, to sum up, I'm pretty sure that the apcupsd slave 'lost connections' are the result of performance/resource issues on the unRAID machine. The 'lost connections' problem returned when I started using docker. So, I would concur that there may be some performance issue with the host services when using docker. One thing I haven't tried is to boot without Xen support. Are there any other experiments that anyone can suggest that I should try in order to verify/track down this issue? An example of the symptoms, as displayed on an xterm session running on the apcupsd slave machine: Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:17:43 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:18:02 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:22:02 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:22:33 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:22:48 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:22:57 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:24:25 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:24:44 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:36:15 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:36:28 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:44:50 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 18:45:03 2014): Communications restored with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 19:36:41 2014): Warning communications lost with UPS Tower Broadcast message from root@desktop (somewhere) (Thu Oct 9 19:36:56 2014): Communications restored with UPS Tower
Archived
This topic is now archived and is closed to further replies.