dmacias Posted September 17, 2016 Share Posted September 17, 2016 ah ok. I haven't had that problem. But I'll look into in now. Lots of posts about it last couple of days... Seems ok for me but I understand. Here's part of my log. This is right after array is started and mounted. Next cache dirs starts then sdh is mounted then docker... Sep 17 09:48:20 server cache_dirs: cache_dirs process ID 16656 started Sep 17 09:48:20 server emhttp: Sep 17 09:48:20 server emhttp: Sep 17 09:48:20 server kernel: XFS (sdh1): Mounting V4 Filesystem Sep 17 09:48:20 server kernel: XFS (sdh1): Ending clean mount Sep 17 09:48:20 server emhttp: Starting services... Sep 17 09:48:20 server emhttp: shcmd (67): set -o pipefail ; /usr/local/sbin/mount_image '/mnt/domains/system/docker/docker.img' /var/lib/docker 20 |& logger Quote Link to comment
drsparks68 Posted September 17, 2016 Share Posted September 17, 2016 Updated from rc5 and since then no WebGUI and no shares are available. Logged in to the console and ran 'diagnostics' (attached). Network.cfg looks correct. At the console, I do see the system trying to ping github: ping: unknown host github.com Can you hit the webUI via the IP address (192.168.0.202)? Try setting static DNS (and gateway addresses) up. 8.8.8.8 / 8.8.4.4 and 192.168.0.1 for a gateway it's already set to static using that IP. WebGUI and shares aren't reachable, seems like networking isn't working at all I added the 8.8.8.8/8.8.4.4 as DNS servers and everything is still unreachable, both ways. I would delete the /flash/config/network.cfg on your flash drive, reboot, and setup your network from scratch. And I can setup networking via the console? No, but unRaid will revert to defaults which should work. If it doesn't then you'll have to manually edit network.cfg and put in the dns / gateway addresses I've already added the dns & gw to the network.cfg; didn't have any effect. # Generated settings: IFNAME[0]="eth0" DESCRIPTION[0]="Public VM Bridge" USE_DHCP[0]="no" IPADDR[0]="192.168.0.202" NETMASK[0]="255.255.255.0" GATEWAY="192.168.0.1" DHCP_KEEPRESOLV="yes" DNS_SERVER1="8.8.8.8" DNS_SERVER2="8.8.4.4" DNS_SERVER3="" MTU[0]="" SYSNICS="1" I removed it, rebooted and checked my router...it's not pulling an IP. If I run ifconfig from the console, eth0 doesn't have an IP. Use the GUI boot option and set up your network. Yeah, after staring at the boot up about 10 times, it finally hit me that I should boot into the GUI. I was able to get the network set up. It works it I set it to DHCP and I can set a static IP, but if I try to use the old IP address (192.168.0.202), it loses connection again. Kinda weird. Now just dealing with Dockers that don't want to start. Quote Link to comment
GerryGER Posted September 17, 2016 Share Posted September 17, 2016 Great to see the 6.2 Stable Release, had been following the Beta/RC development quite a bit but didn't dare to update. Did the update via Web GUI yesterday and it worked flawlessly as far I can tell. Even Auto Sleep works for me again. BUUUT: Is there a more detailed writeup on Turbo Write available than the short text block from the blog (and wiki)? A friend and I are slightly confused on how exactly it's done / works. Why does it spin up all the disks? Does unRAID do striping now? Or does it just distribute files in a different way over the HDDs (ignoring settings like high water/fill up etc.)? How exactly does the new writing algo work? Would be nice to read a bit more about it. Quote Link to comment
jonp Posted September 17, 2016 Share Posted September 17, 2016 Turbo write keeps all disks spinning for a write to any one disk. The net difference is without turbo write, we have to do a read, modify, write. With turbo write it's just a reconstruct write, which is the same method we use to rebuild a failed disk. Quote Link to comment
DoeBoye Posted September 17, 2016 Share Posted September 17, 2016 Upgraded from 6.1.9. Didn't go that smoothly. First had the Dynamix gui bug before the upgrade. Then did the upgrade and system froze. Eventually got it sort of working. Had a hard time with connecting to the web gui. My board has 4 ethernet ports, and something was getting messed up with which port UnRaid wanted to communicate with. Got it working by moving the ethernet cable from one port to the next until I could get the gui to load. Things were still a bit odd, so just backed everything up, wiped the usb key and did a fresh install. Things are running much smoother now. Added a second parity drive and parity is synching as we speak. Now to the crux of the problem: All my shares are showing, but some of the user shares are empty when you click through network shares (Multiple windows machines as well as an ubuntu-based Kodi system), as well as through Putty and Midnight Commander. Disk shares show all files correctly. Before I go through the whole diagnostics posting etc, any obvious things I may have overlooked? I didn't manually create any of the user shares. UnRaid created them from the top level folders on the drives. Could that be related? Also, the really odd thing is that some of the User Shares are fine... Thanks! Quote Link to comment
Glimmerman911 Posted September 17, 2016 Share Posted September 17, 2016 I am trying to upgrade from 6.1.2 to 6.2, but when I check for updates, nothing comes back, it says it is up to date. How do I upgrade? I also have PlexMediaServer from Needo running as a docker, do I need to do anything for this as part of the upgrade? Quote Link to comment
isvein Posted September 17, 2016 Share Posted September 17, 2016 I have problems with docker after upgrade, suddenly the docker image is filled up (20GB) and I dont know why, but I think it may have to do something with Plex. Quote Link to comment
ufopinball Posted September 17, 2016 Share Posted September 17, 2016 Smooth upgrade from 6.1.9 to 6.2 Stable. Quick spot checking shows Plex is working, and my LAMP server in a VM is fine. Will check the other VMs later (some Linux, some Windows). Started a parity check overnight (I juggled some disks) and that completed in 32953sec ... slightly better than average. Looking good... Have now added the 2nd parity disk and that build is running. Watching the Dashboard, the build seems to be using 3 out of 6 cores, easily averaging below 30%, and a max usage below 50%. Not bad for this (Cortex) somewhat older AMD system. Thanks LimeTech, will report back later, but am not presently expecting any issues. Quote Link to comment
Squid Posted September 17, 2016 Share Posted September 17, 2016 I am trying to upgrade from 6.1.2 to 6.2, but when I check for updates, nothing comes back, it says it is up to date. How do I upgrade? I also have PlexMediaServer from Needo running as a docker, do I need to do anything for this as part of the upgrade? You can always download the 6.2 release from LT's website, open up the zip file and extract the files onto your flash drive Quote Link to comment
Glimmerman911 Posted September 17, 2016 Share Posted September 17, 2016 This is the process I have always used, and the release notes would specify which folders & files to copy, and which to leave alone. Which folders and files do I copy? Quote Link to comment
Squid Posted September 17, 2016 Share Posted September 17, 2016 This is the process I have always used, and the release notes would specify which folders & files to copy, and which to leave alone. Which folders and files do I copy? bzroot, bzimage, bzimage-gui, and the syslinux folder. Quote Link to comment
RobJ Posted September 17, 2016 Share Posted September 17, 2016 This is the process I have always used, and the release notes would specify which folders & files to copy, and which to leave alone. Which folders and files do I copy? bzroot, bzimage, bzimage-gui, and the syslinux folder. That's the basic requirements. For more info, see this post. It's now mentioned in the Upgrade Advice post, in the 'not enough RAM' section. Quote Link to comment
bonienl Posted September 17, 2016 Share Posted September 17, 2016 I am trying to upgrade from 6.1.2 to 6.2, but when I check for updates, nothing comes back, it says it is up to date. I installed 6.1.2 on my test server and after a "Check for Updates" the new unRAID version 6.2 is available as update. Quote Link to comment
Squid Posted September 17, 2016 Share Posted September 17, 2016 I am trying to upgrade from 6.1.2 to 6.2, but when I check for updates, nothing comes back, it says it is up to date. I installed 6.1.2 on my test server and after a "Check for Updates" the new unRAID version 6.2 is available as update. Probably a DNS issue then. Setting static DNS addresses and ensuring the gateway is set correctly would fix the problem. Quote Link to comment
garycase Posted September 17, 2016 Share Posted September 17, 2016 I am trying to upgrade from 6.1.2 to 6.2, but when I check for updates, nothing comes back, it says it is up to date. I installed 6.1.2 on my test server and after a "Check for Updates" the new unRAID version 6.2 is available as update. I don't recall exactly when it happened, but at one point there was a line you had to manually add in the Install Plugin section (a URL) and install to add the functionality of checking for updates. If that plugin is already on a system (perhaps bonienl's test server), then it may still function, even if you revert to an earlier version (e.g. 6.1.2). Bonienl => (a) any chance this is what happened?, and (b) do you remember what that plugin is? Quote Link to comment
garycase Posted September 17, 2016 Share Posted September 17, 2016 ... found what I was thinking of => it was way back at 6.0 Beta 12 => so clearly that's NOT the issue !! I suspect Squid's right r.e. a DNS issue => so resolving that should fix the update problem. You could also simply download 6.2 manually and copy the files to your flash drive. Quote Link to comment
bonienl Posted September 17, 2016 Share Posted September 17, 2016 ... found what I was thinking of => it was way back at 6.0 Beta 12 => so clearly that's NOT the issue !! I suspect Squid's right r.e. a DNS issue => so resolving that should fix the update problem. You could also simply download 6.2 manually and copy the files to your flash drive. Are you referring to this "band-aid" plugin, which was used to reactivate notifications (this was a long time ago)? FYI, did a clean install of 6.1.2 without any plugins, and update is given as expected. Quote Link to comment
isvein Posted September 17, 2016 Share Posted September 17, 2016 I have problems with docker after upgrade, suddenly the docker image is filled up (20GB) and I dont know why, but I think it may have to do something with Plex. Found it out. It was crashplan. For some reason crashplan had set incoming backups to an folder inside the docker image even if the mappings is right on the docker settings. Changed this in crashplan gui and now it looks right. Quote Link to comment
Frank1940 Posted September 17, 2016 Share Posted September 17, 2016 A suggestion for Tom ... Good idea. Please use the existing post #2 for this purpose. Thank you Tom! I've made a start, now need moderators and Lime Technology staff to keep it up to date and complete. Users, before you upgrade, please read the first post! Then review the second. I believe you are more likely to have a better upgrade experience if you do read both. Another suggestion for post #2: All users should now go to "Settings" >> "Notification Settings" and set it up so that they will receive notifications of when a problem develops. At first, it probably best to request daily checks and e-mails. Weekly notifications may be less intrusive but monthly notifications are probably too far apart. This action could alert the user to an easily rectified problem so that it can be addressed before multiple problems accumulate that could easily result in data loss! Quote Link to comment
RobJ Posted September 17, 2016 Share Posted September 17, 2016 I've upgraded now too, without any issues. I do have very interesting news about performance though. Here's my recent Parity Check History: 2016-09-16, 05:27:02 21 hr, 19 min, 2 sec 65.2 MB/s OK 6.2-rc4 10 2048,1024,192 2016-09-15, 02:43:39 14 hr, 57 min, 52 sec 92.8 MB/s OK 6.2-rc4 10 2048,1024,994 2016-09-02, 19:52:05 21 hr, 9 min, 51 sec 65.6 MB/s OK 6.2-rc4 10 2048,1024,192 Jun, 02, 00:56:33 14 hr, 46 min, 22 sec 94.0 MB/s OK 6.1.9 11 2048,1024,NA Apr, 04, 16:20:39 15 hr, 50 min, 38 sec 87.7 MB/s OK 6.1.9 11 2048,1024,NA Feb, 22, 15:07:35, (5TB) 14 hr, 37 min, 51 sec 94.9 MB/s OK 6.1.6 10 2048,1024,NA Dec, 10, 02:02:51, (5TB) 14 hr, 7 min, 26 sec 98.4 MB/s OK 6.1.6 10 2048,1024,NA With 6.1.9, my parity checks were about 15 hours. My first parity check with 6.2-rc4 was over 21 hours, an enormous drop in performance. Johnnieblack looked at my tunables and suggested I try 994 for the tunable md_sync_thresh, and you'll see the performance was back to 6.1 levels, at 15 hours. To confirm under identical conditions, I immediately changed md_sync_thresh back to its default of 192 and started another parity check, and you can see the results above. (Second to last column is number of drives, no real effect.) Two huge takeaways from this - one is that in 6.2 tunables really matter(!), and the second is that the relations between the numbers may be more important than the numbers themselves (e.g. md_sync_thresh should be about 30 lower than md_sync_window). The implication here is that those users (such as myself) who have significantly increased their tunable numbers may see a large drop in performance with the default md_sync_thresh of 192. I am going to leave it to johnnieblack, pauven, and Tom of course, to provide guidelines on what those numbers should be, and how they should relate to each other. I believe pauven is coming out with a new version of his tunables tester. I certainly plan to test further, as I really don't know what's best. For now, any user reporting poor disk performance with 6.2 should check their tunables. Once good guidelines have been empirically determined, it would be good to see the corresponding Help updated for these tunables, and perhaps automatic tunable adjustment, according to the best rules. Quote Link to comment
garycase Posted September 17, 2016 Share Posted September 17, 2016 Definitely agree r.e. the tunables. Ever since Pauven provided that great tool to help find the optimal values, I've had notably better performance on all 4 of my servers. It CAN (and as you've clearly shown DOES) make a very notable difference. Quote Link to comment
RobJ Posted September 17, 2016 Share Posted September 17, 2016 Another suggestion for post #2: All users should now go to "Settings" >> "Notification Settings" and set it up so that they will receive notifications of when a problem develops. At first, it probably best to request daily checks and e-mails. Weekly notifications may be less intrusive but monthly notifications are probably too far apart. This action could alert the user to an easily rectified problem so that it can be addressed before multiple problems accumulate that could easily result in data loss! An excellent suggestion! Done! Quote Link to comment
bonienl Posted September 17, 2016 Share Posted September 17, 2016 An excellent suggestion! Done! Perhaps you want to add the additional info for network configuration too (see PM)? Quote Link to comment
RobJ Posted September 17, 2016 Share Posted September 17, 2016 An excellent suggestion! Done! Perhaps you want to add the additional info for network configuration too (see PM)? Also done now. I've been trying to work on one thing at a time, and save, in case anyone else wanted to make changes to the post. I don't know if the way I added the network info is the best way to do it (excellent stuff by the way, and excellence requires credit!). If you would prefer it some other way, just let me know. Actually, you and Squid should probably be moderators! Quote Link to comment
bonienl Posted September 17, 2016 Share Posted September 17, 2016 An excellent suggestion! Done! Perhaps you want to add the additional info for network configuration too (see PM)? Also done now. I've been trying to work on one thing at a time, and save, in case anyone else wanted to make changes to the post. I don't know if the way I added the network info is the best way to do it (excellent stuff by the way, and excellence requires credit!). If you would prefer it some other way, just let me know. Actually, you and Squid should probably be moderators! For 'easier' reading I would remove the quotes and present it as ordinary text, inline with the other sections. Quote Link to comment
Recommended Posts
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.