jumperalex Posted June 6, 2015 Share Posted June 6, 2015 There are different parts dynamically updated: GUI The updating/refreshing of the GUI happens at the frequency set for Page update frequency, default is real-time. Whenever the GUI is refreshed it will read all stored values of emhttp of that moment. emhttp values The emhttp values are updated in the background by the monitor process or in the foreground when the user changes a page in the GUI. The background updating is done every minute. disk temperatures Disk temperatures are handled separately. smartctl is called at the frequency set for Tunable (poll_attributes), default is 30 minutes. The reason this is done at a much lower frequency is to minimize possible interruption of normal disk operation while reading the disk temperature. I love it when something vague and mysterious is clarified. Thanks! This should be in the wiki! There, fixed that for ya ;-) Link to comment
SCSI Posted June 6, 2015 Share Posted June 6, 2015 Successfully upgraded here. Any chance of updating Midnight Commander to the latest version on future releases? Been waiting for Midnight Commander to add a recursive option for chmod and chown and I think they did in later versions as per https://www.midnight-commander.org/ticket/3109 I know we have the command line with the -R option but it is a great addition to MC. Thanks! Link to comment
doorunrun Posted June 6, 2015 Share Posted June 6, 2015 Any reason my dockers cannot reach a DNS server but my unraid can? It's a known issue. From the Defect Reports message group: http://lime-technology.com/forum/index.php?topic=39597.msg380342#msg380342 The quick fix is to set networking to host in the Docker container's config page, save settings and then revert back to bridge mode and save it. Link to comment
Darqfallen Posted June 6, 2015 Share Posted June 6, 2015 It's a known issue. From the Defect Reports message group: http://lime-technology.com/forum/index.php?topic=39597.msg380342#msg380342 The quick fix is to set networking to host in the Docker container's config page, save settings and then revert back to bridge mode and save it. Thanks! Link to comment
RobJ Posted June 6, 2015 Share Posted June 6, 2015 Incidentally, when I went back to the Main screen (I can't say for sure the setting change is the cause), there were temps showing for all but one drive, yet all showed grey balls, spun down. I tested all of them with "hdparm -C", and all but one were "active/idle", the true spun down drive was "standby". There must be conditions where the displayed spin status becomes out of sync with the true spin status. After several hours, long after they should have spun down, they were still showing temps that were drifting up an down. I brought it back in sync by clicking the Spin Up button. There are different parts dynamically updated: GUI The updating/refreshing of the GUI happens at the frequency set for Page update frequency, default is real-time. Whenever the GUI is refreshed it will read all stored values of emhttp of that moment. emhttp values The emhttp values are updated in the background by the monitor process or in the foreground when the user changes a page in the GUI. The background updating is done every minute. disk temperatures Disk temperatures are handled separately. smartctl is called at the frequency set for Tunable (poll_attributes), default is 30 minutes. The reason this is done at a much lower frequency is to minimize possible interruption of normal disk operation while reading the disk temperature. If you put your server to sleep the status will definitely be out of sync when it wakes preventing the server from sleeping again until a disk is spun up. Thank you very much for the clarification, it explains a lot. However, there appears to be a state where the drive spin status is considered 'spun down', yet the drive is spinning, and temps are showing and being updated. That's what happened in both my and reggie's cases. Once in this state, it appears the 'monitor' process(?) won't check the spin status because it *thinks* it can't, because it *thinks* the drive is not spinning? But whatever process is obtaining the temps could either reset the spin status, or notify the 'monitor' that the drive must be spinning. It would be nice if one of the processes above (like monitor?) would just use 'hdparm -C' to check, but perhaps that command is problematic with Areca's and other non-standard controllers? Just trying to get everything back in sync (displayed status with true status) ... Link to comment
Edrikk Posted June 6, 2015 Share Posted June 6, 2015 Please excuse the very dumb question: If the VMs are set to use br0 and get an IP address from the router, why is it that I can't VNC to that IP address / appropriate port? I have to VNC to the UNRAID IP to have VNC connect which is weird.... Link to comment
itimpi Posted June 6, 2015 Share Posted June 6, 2015 Please excuse the very dumb question: If the VMs are set to use br0 and get an IP address from the router, why is it that I can't VNC to that IP address / appropriate port? I have to VNC to the UNRAID IP to have VNC connect which is weird.... you need to think where the VNC server that you are connecting to is actually running. You could VNC to the VM's IP if you installed a VNC server inside the VM itself. However that would only be active when the OS inside the VM had successfully loaded and started the VNC server. What you are doing instead is taking advantage of the fact that the graphics device emulator that is part of KVM running on unRAID supports VNC. You therefore need to use the IP address of where KVM is running which is the UnRAID server address - not the VM address. It is because you are going in at that level that you can see things like the BIOS messages that would normally appear on a PC's screen. Link to comment
Edrikk Posted June 6, 2015 Share Posted June 6, 2015 Makes absolute sense! Thank you! Link to comment
suleimant Posted June 7, 2015 Share Posted June 7, 2015 Upgraded from rc3 to rc4 and VM's now not loading. Moved back to rc3 and issue disappears. Let me know if there are any othert tests I can do. Relevant part (I think) of the log file: Jun 7 10:00:21 T emhttp: Starting libvirt... Jun 7 10:00:21 T logger: Starting libvirtd... Jun 7 10:00:21 T kernel: tun: Universal TUN/TAP device driver, 1.6 Jun 7 10:00:21 T kernel: tun: © 1999-2004 Max Krasnyansky Jun 7 10:00:22 T kernel: Ebtables v2.0 registered Jun 7 10:00:22 T kernel: device virbr0-nic entered promiscuous mode Jun 7 10:00:22 T avahi-daemon[10153]: Joining mDNS multicast group on interface virbr0.IPv4 with address 192.168.122.1. Jun 7 10:00:22 T avahi-daemon[10153]: New relevant interface virbr0.IPv4 for mDNS. Jun 7 10:00:22 T avahi-daemon[10153]: Registering new address record for 192.168.122.1 on virbr0.IPv4. Jun 7 10:00:22 T kernel: virbr0: port 1(virbr0-nic) entered listening state Jun 7 10:00:22 T kernel: virbr0: port 1(virbr0-nic) entered listening state Jun 7 10:00:22 T dnsmasq[10776]: started, version 2.57 cachesize 150 Jun 7 10:00:22 T dnsmasq[10776]: compile time options: IPv6 GNU-getopt no-DBus I18N DHCP TFTP IDN Jun 7 10:00:22 T dnsmasq-dhcp[10776]: DHCP, IP range 192.168.122.2 -- 192.168.122.254, lease time 1h Jun 7 10:00:22 T dnsmasq[10776]: reading /etc/resolv.conf Jun 7 10:00:22 T dnsmasq[10776]: using nameserver 192.168.1.1#53 Jun 7 10:00:22 T dnsmasq[10776]: read /etc/hosts - 1 addresses Jun 7 10:00:22 T dnsmasq[10776]: read /var/lib/libvirt/dnsmasq/default.addnhosts - 0 addresses Jun 7 10:00:22 T dnsmasq-dhcp[10776]: read /var/lib/libvirt/dnsmasq/default.hostsfile Jun 7 10:00:22 T kernel: virbr0: port 1(virbr0-nic) entered disabled state Jun 7 10:00:23 T ntpd[1492]: Listen normally on 4 virbr0 192.168.122.1:123 Jun 7 10:00:23 T ntpd[1492]: new interface(s) found: waking up resolver Jun 7 10:00:33 T kernel: docker0: port 1(veth220da9e) entered forwarding state Jun 7 10:00:33 T kernel: docker0: port 2(veth046e65d) entered forwarding state Jun 7 10:00:34 T kernel: docker0: port 3(vethbe3047d) entered forwarding state Jun 7 10:00:34 T kernel: br0: topology change detected, propagating Jun 7 10:00:34 T kernel: br0: port 1(eth0) entered forwarding state Jun 7 10:00:34 T kernel: docker0: port 4(vethcdf1f8c) entered forwarding state Jun 7 10:00:45 T kernel: kvm: already loaded the other module Jun 7 10:00:45 T kernel: php[11589]: segfault at 0 ip 00002b71b16540fa sp 00007ffdafd9b290 error 4 in libvirt-php.so[2b71b1642000+36000] Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 Upgrade to rc4 was seamless and easy. Wonderful upgrade facility in v6. I am almost speechless. Only 1 issue seen so far: I don't know where this issue is either rc4 or Community Repositories Plugin but since I upgraded to rc4 the images over the "Click Add to Container" link on Docker Community Repository link is not showing. Either the image has been deleted, a permission issue or something I haven't thought of. Clearly rc4 update wouldn't have affected he plugin itself (or I wouldn't have thought so). Checked this in 2 browsers (Safari and Chrome) and it is not a caching issue as I cleared it on both browsers and reloaded. File 1 is Safari File 2 is Chrome Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 Upgrade to rc4 was seamless and easy. Wonderful upgrade facility in v6. I am almost speechless. Only 1 issue seen so far: I don't know where this issue is either rc4 or Community Repositories Plugin but since I upgraded to rc4 the images over the "Click Add to Container" link on Docker Community Repository link is not showing. Either the image has been deleted, a permission issue or something I haven't thought of. Clearly rc4 update wouldn't have affected he plugin itself (or I wouldn't have thought so). Checked this in 2 browsers (Safari and Chrome) and it is not a caching issue as I cleared it on both browsers and reloaded. File 1 is Safari File 2 is Chrome 2 things: firstly the community repositories plugin is now deprecated and has been replaced with community applications (although this has nothing to do with your issue) Secondly, it appears that your issue is the local caching issue that affects both of the plugins on some systems. A fix is in community applications (look in the settings) Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 Upgrade to rc4 was seamless and easy. Wonderful upgrade facility in v6. I am almost speechless. Only 1 issue seen so far: I don't know where this issue is either rc4 or Community Repositories Plugin but since I upgraded to rc4 the images over the "Click Add to Container" link on Docker Community Repository link is not showing. Either the image has been deleted, a permission issue or something I haven't thought of. Clearly rc4 update wouldn't have affected he plugin itself (or I wouldn't have thought so). Checked this in 2 browsers (Safari and Chrome) and it is not a caching issue as I cleared it on both browsers and reloaded. File 1 is Safari File 2 is Chrome 2 things: firstly the community repositories plugin is now deprecated and has been replaced with community applications (although this has nothing to do with your issue) Secondly, it appears that your issue is the local caching issue that affects both of the plugins on some systems. A fix is in community applications (look in the settings) Wow - thanks man! I did not know that. I shall remove the plugin and install the new one. I am going to re-itterate a request I made some time ago, there NEEDS to be some communication system with Plugins / Dockers. If there has been depreciation it CANNOT only be known IF someone stumbles on the correct post in the forum. Surely it would be easy enough to add an RSS feed or similar as an extra column in the plugin table and / or some notification in unraid or simply an email when something like this happens! Even the most basic of service has an notification system!! Anyway that was for LT et al - @Squid - Thanks man! Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 Upgrade to rc4 was seamless and easy. Wonderful upgrade facility in v6. I am almost speechless. Only 1 issue seen so far: I don't know where this issue is either rc4 or Community Repositories Plugin but since I upgraded to rc4 the images over the "Click Add to Container" link on Docker Community Repository link is not showing. Either the image has been deleted, a permission issue or something I haven't thought of. Clearly rc4 update wouldn't have affected he plugin itself (or I wouldn't have thought so). Checked this in 2 browsers (Safari and Chrome) and it is not a caching issue as I cleared it on both browsers and reloaded. File 1 is Safari File 2 is Chrome 2 things: firstly the community repositories plugin is now deprecated and has been replaced with community applications (although this has nothing to do with your issue) Secondly, it appears that your issue is the local caching issue that affects both of the plugins on some systems. A fix is in community applications (look in the settings) Wow - thanks man! I did not know that. I shall remove the plugin and install the new one. I am going to re-itterate a request I made some time ago, there NEEDS to be some communication system with Plugins / Dockers. If there has been depreciation it CANNOT only be known IF someone stumbles on the correct post in the forum. Surely it would be easy enough to add an RSS feed or similar as an extra column and / or some notification in unraid or simply and email when something like this happens! Even the most basic of service has an notification system!! Anyway that was for LT et al - @Squid - Thanks man! I had already thought of that and wrote an update for community repositories that would pop up a warning that it was deprecated, but ultimately decided not to release it for a while because the plugin itself is still working correctly. And I thought I might get some backlash from the community for doing that (since it would be a first). But I still might do that when 6.0 goes final Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 Upgrade to rc4 was seamless and easy. Wonderful upgrade facility in v6. I am almost speechless. Only 1 issue seen so far: I don't know where this issue is either rc4 or Community Repositories Plugin but since I upgraded to rc4 the images over the "Click Add to Container" link on Docker Community Repository link is not showing. Either the image has been deleted, a permission issue or something I haven't thought of. Clearly rc4 update wouldn't have affected he plugin itself (or I wouldn't have thought so). Checked this in 2 browsers (Safari and Chrome) and it is not a caching issue as I cleared it on both browsers and reloaded. File 1 is Safari File 2 is Chrome 2 things: firstly the community repositories plugin is now deprecated and has been replaced with community applications (although this has nothing to do with your issue) Secondly, it appears that your issue is the local caching issue that affects both of the plugins on some systems. A fix is in community applications (look in the settings) Wow - thanks man! I did not know that. I shall remove the plugin and install the new one. I am going to re-itterate a request I made some time ago, there NEEDS to be some communication system with Plugins / Dockers. If there has been depreciation it CANNOT only be known IF someone stumbles on the correct post in the forum. Surely it would be easy enough to add an RSS feed or similar as an extra column and / or some notification in unraid or simply and email when something like this happens! Even the most basic of service has an notification system!! Anyway that was for LT et al - @Squid - Thanks man! I had already thought of that and wrote an update for community repositories that would pop up a warning that it was deprecated, but ultimately decided not to release it for a while because the plugin itself is still working correctly. And I thought I might get some backlash from the community for doing that (since it would be a first). But I still might do that when 6.0 goes final That would get a HUGE +1 from me. Perhaps LT need to create a centralised notification system with an api you can call But I guess like lots of other features its in the future. I have "tone" in this post because quite frankly I am pissed! I have just spent the best part of this long weekend here in ViC Australia getting my Unraid server working great (yes using community repositories plugin) and have things installed like: Sabnzbd (needo) Sickrage (needo) Couchpotato (needo) owncloud (gfjardim) HTPC Manager (smdion) Kodi headless (sparklyballs) Maraschino (smdion) MariaDB (needo) ReverseProxy (smdion) NOW with the "NEW" Community Applications Plugin when I search for these Dockers I get versions created and maintained by others. So I am guessing that these are now the "Recommended" versions I should be installing and not the ones I have spent hours configuring and getting right. For instance apps created by needo doesn't show up as a repository anymore!? Now I seem to be directed at binhex and hurricane. But which one to select? And what happened to needo ones I JUST installed a matter of hours ago!??! Same with owncloud, I was JUST purring over it and how nice it is. I JUST installed the version created by gfjardim and NOW when I click "Cloud" in the new Community Application Plugin and I am directed to a version created by l3iggs in Bungy's repo!!!??!! This has the potential to get WAY out of hand. And for me it already has. I almost feel the need to change my versions to those "recommend" in the Community Application Plugin because one would feel it gets the most support. Amazing how you can go from Happy to Disappointed in a matter of hours. EDIT: OK - perhaps I posted too soon. It seems those I have installed are still there but almost ALL of them are in the Uncategorised section. One must ask WHY they are uncategorised?? Does this MEAN something - are they about to be depreciated too? WHY is there even an uncategorised section? Are the ones in the categorised section "better" than the ones in the Uncategorised section? Sigh. Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 Don't get upset. Categorization is only 2 weeks old, and the release of community apps coincided with gfjatdim going on vacation so those applications from him and needo have not been categorized yet. (They still show up under the uncategorized category and I am sure that they will be categorized shortly). For what its worth I personally use needo's version also Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 Don't get upset. Categorization is only 2 weeks old, and the release of community apps coincided with gfjatdim going on vacation so those applications from him and needo have not been categorized yet. (They still show up under the uncategorized category and I am sure that they will be categorized shortly). For what its worth I personally use needo's version also Thanks for that post. Very patient person. I might have to amend my last post just in case it comes across "below the line". I've taken a breath (and a nip of Jim Beam Small Batch) and need to remember this is all a WIP. I think think the notification modification you mentioned you have been working on is a MUST. Especially with the fluidity of things at the moment, If you need a tester or helper - you have one right here. I have 3 systems (Main, Backup and Test) at your disposal for such an important feature. Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 What I was working on wasn't a general notification system per se. Rather it was a specific update for a plugin that took advantage of the fact the v6 already has a notification system in place to alert users of plugin updates to let the user know that the plugin was deprecated. And it was more of a proof of concept for another idea. Link to comment
bonienl Posted June 7, 2015 Share Posted June 7, 2015 I think think the notification modification you mentioned you have been working on is a MUST. Especially with the fluidity of things at the moment, If you need a tester or helper - you have one right here. I have 3 systems (Main, Backup and Test) at your disposal for such an important feature. The current notification system allows developers to write CHANGES, for example - a particular plugin is depreciated, the developer can issue a 'new version' with an announcement and point to the substitute version. When people have the notification system set up to check for updates, they will get notified and subsequently can check the changes and do the appropriate steps. Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 I think think the notification modification you mentioned you have been working on is a MUST. Especially with the fluidity of things at the moment, If you need a tester or helper - you have one right here. I have 3 systems (Main, Backup and Test) at your disposal for such an important feature. The current notification system allows developers to write CHANGES, for example - a particular plugin is depreciated, the developer can issue a 'new version' with an announcement and point to the substitute version. When people have the notification system set up to check for updates, they will get notified and subsequently can check the changes and do the appropriate steps. Thank you for that post. So - if I am reading your post correctly, in the "Community Repositories Plugin" example - the developer just didn't post a new version indicating that it was depreciated or I didn't have the system set up to notify me or both? How does one setup the system to notify me automatically if there is a change? Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 I think think the notification modification you mentioned you have been working on is a MUST. Especially with the fluidity of things at the moment, If you need a tester or helper - you have one right here. I have 3 systems (Main, Backup and Test) at your disposal for such an important feature. The current notification system allows developers to write CHANGES, for example - a particular plugin is depreciated, the developer can issue a 'new version' with an announcement and point to the substitute version. When people have the notification system set up to check for updates, they will get notified and subsequently can check the changes and do the appropriate steps. Thank you for that post. So - if I am reading your post correctly, in the "Community Repositories Plugin" example - the developer just didn't post a new version indicating that it was depreciated or I didn't have the system set up to notify me or both? How does one setup the system to notify me automatically if there is a change? As far as community repositories is concerned, I only officially deprecated it a few days ago. I never bothered to release an update for the plugin to alert users of this because fundamentally at its core community applications is the same plugin just with reorganized information. I could have just as easily pushed out community applications as an update to community repositories. Also, just because the plugin is deprecated does not mean that it does not work. It merely means that no further development is being done on it. I actually went to great pains to ensure that the data which both plugins collect about the repositories / applications is compatible with each other. Also just as a note / disclaimer: none of the applications which are displayed by either plugin are recommended by either myself or Limetech. All applications have been created by the community forum and are supported within the forum. Link to comment
sparklyballs Posted June 7, 2015 Share Posted June 7, 2015 i'd like to find the person responsible for suggesting categories and give him a piece of my mind, lol. Link to comment
danioj Posted June 7, 2015 Share Posted June 7, 2015 I think think the notification modification you mentioned you have been working on is a MUST. Especially with the fluidity of things at the moment, If you need a tester or helper - you have one right here. I have 3 systems (Main, Backup and Test) at your disposal for such an important feature. The current notification system allows developers to write CHANGES, for example - a particular plugin is depreciated, the developer can issue a 'new version' with an announcement and point to the substitute version. When people have the notification system set up to check for updates, they will get notified and subsequently can check the changes and do the appropriate steps. Thank you for that post. So - if I am reading your post correctly, in the "Community Repositories Plugin" example - the developer just didn't post a new version indicating that it was depreciated or I didn't have the system set up to notify me or both? How does one setup the system to notify me automatically if there is a change? As far as community repositories is concerned, I only officially deprecated it a few days ago. I never bothered to release an update for the plugin to alert users of this because fundamentally at its core community applications is the same plugin just with reorganized information. I could have just as easily pushed out community applications as an update to community repositories. Also, just because the plugin is deprecated does not mean that it does not work. It merely means that no further development is being done on it. I actually went to great pains to ensure that the data which both plugins collect about the repositories / applications is compatible with each other. Also just as a note / disclaimer: none of the applications which are displayed by either plugin are recommended by either myself or Limetech. All applications have been created by the community forum and are supported within the forum. I am sorry that you felt the need to publish an explanation / disclaimer for all your hard work that we are all benefiting from. If it is any consolation I did not even know who was responsible for either of the plugin's when I was posting. The issue is systemic and not based on the hard work of individuals. If we come up with an iron clad process which ALL need to follow then we will eliminate this issue in the future. Lets keep this clean and not a mess. Wouldn't it be nice if there was a 1:1 between Unraid Docker and Application and not 1:3/x. So if I want application x I don't have to select from 1 or 3 (or more) possible options. I do feel LT has a vested interest in such a process as they "market" their product using the features of Docker heavily nowadays - wouldn't it be nice to keep it clean! @Squid: I for one would like to take the time to publically appreciate your dedication, support, time taken to post and general effort in what you are doing. Link to comment
HellDiverUK Posted June 7, 2015 Share Posted June 7, 2015 Just like to say I've got RC4 up and running on my X10SBA. Running perfectly so far. Link to comment
Squid Posted June 7, 2015 Share Posted June 7, 2015 You're welcome. The reason for my "disclaimer" was that a couple of posts up you referred to "recommended" applications. Thought that I should clear it up before since I don't think it had ever been mentioned before. Personally, I believe that choice amongst the applications is a good thing, however I would like to see the authors take advantage of the tools which the plugins give them to show the users what the differences between binhex versions and needo versions. Link to comment
bonienl Posted June 7, 2015 Share Posted June 7, 2015 At the moment the Community Applications plugin is the hard work of Squid, and we should be thankful to him to have started this initiative and he goes into great lengths to keep it up to date. The usefulness of this plugin is undoubtably there and surely the potential exists to integrate it with a future version of unRAID. All you need to have is patience and appreciate what we have today. Link to comment
Recommended Posts