August 27, 201411 yr Hence why I added my syslog in the hope someone else could help me. The syslog doesn't show any other issues. However, this and the previous syslog were NOT saved by Powerdown, so there is so far no evidence that Powerdown 2.06a is running. How do you reboot? When you view the console while it reboots, do you see the "Powerdown V2.06a" or "Powerdown V2.06"? If Disk Utility is not fixing to flash then something else is wrong. After running Disk Utility on the flash and properly ejecting the flash from the Mac, the next boot of the system should not have the message. As he says, there is no way that message should appear in a syslog immediately after you have fixed the flash drive. Something here is wrong. Can you give us the exact steps you are taking? A side note, I created a Powerdown 2.07, but Dynamix appears to have Powerdown 2.06 hardwired (the name powerdown-2.06-noarch-unRAID.tgz is hard coded in it), so 2.07 was 'upgraded' to 2.06. Perhaps renaming it as the former would work, but I find that an unacceptable practice. In addition, I discovered that my powerdown-2.06-noarch-unRAID.tgz is not the same as the one currently downloading from the git repository. Mine has files dated Mar 1, 2014, and the current one has files dated Mar 4, with 2 important files within having been revised. The revisions aren't major, but whoever did that (making code changes but leaving the names and version numbers identical), consider this a slap on the hand!
August 27, 201411 yr Seems like we are getting somewhere. So Dynamix could be causing me the problems. Here is a screen shot that mentions your name but it looks like its still on 2.06. Its pointless booting in safe mode because I need the APC UPS script and powerdown. I found the same when adding cache dirs from Dynamix. I don't know what is going on but people seem to be changing plugins all the time and it drives me crazy if revision numbers are not in there. I added Joe's Cache Dirs script outside of Dynamix. To be honest I do really like Dynamix especially in the back theme but hardcoding someone else's script into it doesn't really make sense.
August 27, 201411 yr Seems like we are getting somewhere. So Dynamix could be causing me the problems. Here is a screen shot that mentions your name but it looks like its still on 2.06. Its pointless booting in safe mode because I need the APC UPS script and powerdown. I found the same when adding cache dirs from Dynamix. I don't know what is going on but people seem to be changing plugins all the time and it drives me crazy if revision numbers are not in there. I added Joe's Cache Dirs script outside of Dynamix. To be honest I do really like Dynamix especially in the back theme but hardcoding someone else's script into it doesn't really make sense. No I don't see a direct connection with Dynamix, that was just a side note to others. The only issue here with Dynamix is that it makes it harder to revise Powerdown. Here is a screen shot that mentions your name While I'm flattered to be confused with such a great person as Robert Cotrone (our beloved WeeboTech), I can't usurp his prestige and fame! My name isn't mentioned anywhere there.
August 27, 201411 yr I didn't realise that was WeeboTech's name and thought it was probably yours. My mistake. Sorry WeeboTech. So I have added you file to my boot directory and also added the command to the go file and still no luck. Is anyone else trying this while using Dynamix or is everyone on the stock WebGui.
August 28, 201411 yr can you provide the screen shot of the last few screens as the machine shuts down. the sync should insure all blocks are written, so perhaps something has the filesystem busy and it cannot be unmounted. As Im using KVM its not easy, any ideas on how I can do this?. I do notice it trys many times to unmount the disks when I stop the array before a power down or reboot. Thanks
August 28, 201411 yr Perhaps we need to put some kind of fuser statement to see what might be holding /boot active. I know we kick (kill) pids on the array, but I don't remember doing that on /boot. I'm grasping at straws here and since I'm very busy and do not have the latest source, I can recommend a line to see what might be . From there we may be able to add /boot as part of the final kill. This like will show what pids are active on /boot. From there, maybe we can determine if the filesystem is being held hostage. for pid in $(fuser -cu /boot 2>/dev/null) ; do ps --no-headers -fp ${pid}; pstree ${pid} ; done here's an example of what the output might look like. root@unRAID:/# for pid in $(fuser -cu /boot 2>/dev/null) ; do ps --no-headers -fp ${pid}; pstree ${pid} ; done Shows nothing on /boot root@unRAID:/# cd /boot root@unRAID:/boot# for pid in $(fuser -cu /boot 2>/dev/null) ; do ps --no-headers -fp ${pid}; pstree ${pid} ; done root 19288 8004 0 23:17 pts/5 00:00:00 -bash bash---pstree Shows my process on /boot. I suppose you can put this after the sync that RobJ added. From there let's see what might be holding up the filesystem. If that doesn't work, maybe the flash card in use requires more time to do it's writes. So adding a sleep after the sync may give more time (lets use a full minute here just to be sure) In any case, Keep an eye out. You might want to do a shutdown and capture the last screen for others to review. Don't do the poweroff, just a regular shutdown. The failure as the filesystem is being mounted doesn't tell us much, but a failure as it's coming down might.
August 28, 201411 yr WeeboTech can you explain this please. "Don't do the poweroff, just a regular shutdown" All I have is the power off in dynamix or when my WebGui hangs I use the powerdown or powerdown -r from command line. Thanks
August 28, 201411 yr I don't know the dymanix front end. I thought the standard unRAID front end had power off and shutdown. The real key is catching the last screen to see what's going on, with out that all of this is useless. When debugging this particular scenario, you should probably be in safe mode as none of us know it's powerdown, a problem with unRAID or the specific flash. I suppose if powerdown was suspect, that should be the only thing installed.
August 28, 201411 yr A side note, I created a Powerdown 2.07, but Dynamix appears to have Powerdown 2.06 hardwired (the name powerdown-2.06-noarch-unRAID.tgz is hard coded in it), so 2.07 was 'upgraded' to 2.06. Perhaps renaming it as the former would work, but I find that an unacceptable practice. In addition, I discovered that my powerdown-2.06-noarch-unRAID.tgz is not the same as the one currently downloading from the git repository. Mine has files dated Mar 1, 2014, and the current one has files dated Mar 4, with 2 important files within having been revised. The revisions aren't major, but whoever did that (making code changes but leaving the names and version numbers identical), consider this a slap on the hand! Dynamix webGui has a reference to Powerdown 2.06 on github in the PLG file. This gets loaded first. If you want to upgrade to a later version (which by the way is not in the github repository), you can install the separate Powerdown PLG either manually or place the PLG file in '/boot/config/plugins' and reboot your system. It *should* upgrade the version to the latest, though proper naming of the PLG file is relevant here.
August 28, 201411 yr WeeboTech, I still get the error in safe mode so I really need to get my hands on a windows PC first so I can clean the SD card. I have captured an AVI file of the reboot and startup but its 43mb in size so I can post it on the forums.
August 28, 201411 yr Author I'm coming to the party a little late so I am not sure I am totally up to speed on the issue here. It seems that a shutdown does not allow the flash drive to be cleanly shutdown and unRAID shows the unclean shutdown message when re-booting? When you boot in the safe mode and do a normal shutdown through the webgui does the problem occur?
August 28, 201411 yr Not to the best of my knowledge although from what other users have said. Im running Dynamix though and have to find a windows PC to run chkdsk on my flash card first as its already giving errors in safe mode and normal with Dynamix, Dynamix System Temp, APC UPS and cache dirs. Only then can I really confirm if safe mode or the stock WebGui gives errors or not and if RobJ's mod will work with Dynamix. I raised the point about noticing the flash error a few weeks back in my KVM at the command line.
August 28, 201411 yr Author I guess I'm a bit confused. I would suggest: - Fix the flash drive with a PC or Mac. - Start unRAID in safe mode and then do a normal webgui shutdown. See if you get the flash error on the next boot. - Remove all plugins except for apcupsd and the start unRAID. See if you get the flash error on the next boot. If you do get the error after installing the apcupsd plugin, post the log from the /logs directory on the flash device. If you've already done these things, just let me know the results and we can take the next step.
August 28, 201411 yr I don't know if RobJ wrote the mod because he was having problems as well as others, perhaps he can shed more light on it. But I will let you know as soon as I have sorted out my SD Card.
August 28, 201411 yr Author I'm thinking that maybe the flash is not getting enough time to complete its writes. I will add a slight time delay after the syslog is written to the flash drive. The workaround "patch" is probably not in the best place. I have a new version of powerdown that I was holding until v6 settled down. I'll add a time delay and then release a version 2.07. Powerdown will probably be replaced in v6 final by the apcupsd that Tom will implement native in unRAID, but until that happens I'll keep Powerdown up to date as best I can. I am very busy at work and don't have a lot of time these days to put towards unRAID projects. One new feature of Powerdown is the ability to control the saving of logs. As I remember you can control the number of logs to keep - the default is 10 with a configuration file parameter.
August 29, 201411 yr Thanks dlandon and thanks to RobJ and WeeboTech for all their help. As I use Dynamix will this be in the git repository that Bonieni spoke about so it updates with Dynamix?
August 29, 201411 yr Author Thanks dlandon and thanks to RobJ and WeeboTech for all their help. As I use Dynamix will this be in the git repository that Bonieni spoke about so it updates with Dynamix? It will be released to the git depository and bonienl will have to modify Dynamix to pick up the new version.
August 29, 201411 yr I have an updated version of Dynamix in the planning, will wait until the new powerdown version is available and include that as well.
August 31, 201411 yr Author Version 2.07 of powerdown is now available. A sync and time delay have been added to the shutdown process to insure that logs can be written to the flash drive and the flash drive complete the write operation. A new feature is the ability to specify the location of historical log files and how many to keep. Refer to the OP for details.
September 3, 201411 yr Author I have created a plugin for powerdown for V6 that is loaded by Apcupsd as a plugin. This exposes the powerdown plugin in the plugin manager so the installation, and update can be managed independently from any plugin that uses powerdown. Because the powerdown plugin controls the installation of powerdown, plugins such as Apcupsd only need to install the powerdown plugin and no longer need to be concerned about versions. I'm expecting other plugin authors to adopt the same scheme and not install powerdown as a package, but as a plugin so installation and version control are better managed and understood by the user.
September 4, 201411 yr Thanks Dlandon and thanks for making the changes to 2.07. I have not have any errors on my flash drive since your update.
September 15, 201411 yr Author There is an issue in the latest V6 beta9 with docker not being stopped when a shut down outside the webgui is initiated. i.e. when the power button is pressed and the array has not been stopped. This is a task for unRAID. The powerdown package should not do this. I am making an adjustment in powerdown because it is stopping the new logger daemon and it should not do that. Logging is being terminated so the final shutdown is not logged. This only causes an issue in the logging. It doesn't have anything to do with an unclean shutdown.
September 15, 201411 yr Author I presume you mean with v6 beta 9? Yes. I'm working with V6b9 so that's where I found it, but I think it's in earlier versions also.
Archived
This topic is now archived and is closed to further replies.