July 16, 201312 yr noticed a syslog message today, 'kernel: Disabling IRQ #16', looking at the syslog its: Jul 15 10:06:41 husky kernel: mdcmd (144): spindown 3 (Routine) Jul 15 10:25:56 husky kernel: irq 16: nobody cared (try booting with the "irqpoll" option) (Errors) Jul 15 10:25:56 husky kernel: Pid: 0, comm: swapper/0 Tainted: G O 3.9.6p-unRAID #23 (Errors) Jul 15 10:25:56 husky kernel: Call Trace: (Errors) Jul 15 10:25:56 husky kernel: [<c105f89e>] __report_bad_irq+0x29/0xb4 (Errors) Jul 15 10:25:56 husky kernel: [<c105fa60>] note_interrupt+0x137/0x1ac (Errors) Jul 15 10:25:56 husky kernel: [<c135d6e1>] ? cpuidle_wrap_enter+0x2f/0x82 (Errors) Jul 15 10:25:56 husky kernel: [<c105dfd0>] handle_irq_event_percpu+0x109/0x11a (Errors) Jul 15 10:25:56 husky kernel: [<c105ff24>] ? handle_percpu_irq+0x3b/0x3b (Errors) Jul 15 10:25:56 husky kernel: [<c105e006>] handle_irq_event+0x25/0x3c (Errors) Jul 15 10:25:56 husky kernel: [<c105ff24>] ? handle_percpu_irq+0x3b/0x3b (Errors) Jul 15 10:25:56 husky kernel: [<c105ff91>] handle_fasteoi_irq+0x6d/0xab (Errors) Jul 15 10:25:56 husky kernel: <IRQ> [<c100362e>] ? do_IRQ+0x37/0x9b Jul 15 10:25:56 husky kernel: [<c1403d2c>] ? common_interrupt+0x2c/0x31 (Errors) Jul 15 10:25:56 husky kernel: [<c135d6e1>] ? cpuidle_wrap_enter+0x2f/0x82 (Errors) Jul 15 10:25:56 husky kernel: [<c135d746>] ? cpuidle_enter_tk+0x12/0x14 (Errors) Jul 15 10:25:56 husky kernel: [<c135d1d1>] ? disable_cpuidle+0xf/0xf (Errors) Jul 15 10:25:56 husky kernel: [<c135d1ef>] ? cpuidle_enter_state+0xc/0x38 (Errors) Jul 15 10:25:56 husky kernel: [<c135d605>] ? cpuidle_idle_call+0x73/0x9b (Errors) Jul 15 10:25:56 husky kernel: [<c1008802>] ? cpu_idle+0x46/0x6f (Errors) Jul 15 10:25:56 husky kernel: [<c13f6b98>] ? rest_init+0x58/0x5a (Errors) Jul 15 10:25:56 husky kernel: [<c1574a0d>] ? start_kernel+0x2ad/0x2b3 (Errors) Jul 15 10:25:56 husky kernel: [<c15745db>] ? repair_env_string+0x53/0x53 (Errors) Jul 15 10:25:56 husky kernel: [<c15742a3>] ? i386_start_kernel+0x79/0x7d (Errors) Jul 15 10:25:56 husky kernel: handlers: Jul 15 10:25:56 husky kernel: [<c1321628>] usb_hcd_irq (Drive related) Jul 15 10:25:56 husky kernel: [<f848009d>] mvs_interrupt [mvsas] (Drive related) Jul 15 10:25:56 husky kernel: Disabling IRQ #16 Jul 15 11:19:43 husky kernel: mdcmd (145): spindown 3 (Routine) anyone seen this before? Only USB devices I have connected is the UPS... just tried and unmenu can still poll it just fine. I did have a USB mouse attached but disconnected it since I didn't even have a keyboard attached anymore... reading through the forums some people say this could be the culprit. Regardless, looking though the 'pci devices' I see two entries for IRQ 16: 00:1a.0 USB Controller: Intel Corporation Unknown device 1e2d (rev 04) (prog-if 20 [EHCI]) Subsystem: Micro-Star International Co., Ltd. Unknown device 7752 Flags: bus master, medium devsel, latency 0, IRQ 16 Memory at f7f13000 (32-bit, non-prefetchable) [size=1K] Capabilities: [50] Power Management version 2 Capabilities: [58] Debug port: BAR=1 offset=00a0 Capabilities: [98] PCIe advanced features Kernel driver in use: ehci-pci 01:00.0 RAID bus controller: Marvell Technology Group Ltd. MV64460/64461/64462 System Controller, Revision B (rev 01) Subsystem: Marvell Technology Group Ltd. Unknown device 6480 Flags: bus master, fast devsel, latency 0, IRQ 16 I/O ports at e000 [size=128] Memory at f7e40000 (64-bit, non-prefetchable) [size=64K] Expansion ROM at f7e00000 [disabled] [size=256K] Capabilities: [48] Power Management version 2 Capabilities: [50] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable- Capabilities: [e0] Express Legacy Endpoint, MSI 00 Capabilities: [100] Advanced Error Reporting Kernel driver in use: mvsas Kernel modules: mvsas so far everything appears fine.. thinking about restarting the box just in case... will wait to hear back from Tom first.
July 17, 201312 yr noticed a syslog message today, 'kernel: Disabling IRQ #16', ... so far everything appears fine.. thinking about restarting the box just in case... will wait to hear back from Tom first. Here is a whole thread on my IRQ #16 being disabled issue. Like with your system, IRQ 16 gets multiple assignments in my system. Simply switching my display input disabled IRQ #16 which killed my Marvell based SATA controller and the I/O to all drives attached to it. lime-technology.com/forum/index.php?topic=17823.0 Don't think this is an RC16c issue.
July 17, 201312 yr wish unraid could warn me about irq's being assigned to critical things and plug-n-play devices.. which could result in disastrous problems from something so harmless.. if someone was to just unplug a mouse/keyboard/switch video/whatever and have half their array drop out. # cat /proc/interrupts CPU0 CPU1 CPU2 CPU3 0: 12 0 0 0 IO-APIC-edge timer 1: 1 0 2 0 IO-APIC-edge i8042 9: 0 0 0 0 IO-APIC-fasteoi acpi 12: 0 0 3 0 IO-APIC-edge i8042 16: 89633 5892 1499972 4725 IO-APIC-fasteoi ehci_hcd:usb1, mvsas 17: 9885 2247 80028 1955 IO-APIC-fasteoi mvsas 23: 504 359 2885 326 IO-APIC-fasteoi ehci_hcd:usb2 42: 0 0 0 0 PCI-MSI-edge xhci_hcd 43: 3676852 1674574 121771747 2378309 PCI-MSI-edge eth0 44: 120789 67740 263688 62826 PCI-MSI-edge ahci NMI: 0 0 0 0 Non-maskable interrupts LOC: 5408092 3146394 4848004 1784953 Local timer interrupts SPU: 0 0 0 0 Spurious interrupts PMI: 0 0 0 0 Performance monitoring interrupts IWI: 2 0 0 0 IRQ work interrupts RTR: 3 0 0 0 APIC ICR read retries RES: 62805 92017 19523 13617 Rescheduling interrupts CAL: 17599 14659 20829 16744 Function call interrupts TLB: 2286 2085 3125 2369 TLB shootdowns TRM: 0 0 0 0 Thermal event interrupts THR: 0 0 0 0 Threshold APIC interrupts MCE: 0 0 0 0 Machine check exceptions MCP: 2153 2153 2153 2153 Machine check polls ERR: 0 MIS: 0 guess moral of the story is just dont touch anything connected to the unraid box once its running..
July 17, 201312 yr Tom when did the unRAID code change (release) and add "fuse_remember="330" to the share.cfg file? only when clicking "apply" in the smb settings page? Doing so "clicking apply in the smb settings page, crashed emhttp on my system, services restarted and shares/disk accessible but WebGui inaccessible). I don't see any statement in the release notes to add or click apply to have this additional value set in share.cfg file. Please capture system log. The presence or absence of this setting should not cause any crashes. Syslog attached for you (managed to grab that before my un-graceful restart). When I hit apply in the smb settings, I was not aware it would re-start smb as I did not change any settings/values . Just wanted to see if it would write this new entry as it did for that other person in that post. I had a ISO extracting on the cache drive at the time, when I looked at my ssh session with a tail of syslog, I noticed umount was not successfully (not sure what was unmounting??? to restart smb), so I quickly, canceled the ISO extract, which bombed due to this smb restart and just exited out of the app that was doing the extract from a windows client, the syslog entries proceeded after exiting the app (just letting you know some history to apply while reading the syslog attached, there was no app installed on unRAID nor do I have any installed on unRAID to be clear). Upon completion of smb restart from clicking apply, the WebGui was no longer available, I tried from several different clients. ssh/disk/user share via SMB were available, didn't have time to check the mac clients via AFP. I can understand the presence or absence of this setting, should not have, but emhttp definitely did die when clicking "apply" in the smb settings (whether or not it was do to writing this value). The server was not being taxed, all drives were spun down less the cache drive where the source iso resided and a windows client was extracting the iso in the same folder as the iso. Nothing else was going on, all mac's were asleep. Secondary, so how does one reboot gracefully via command line, as I need to right now. Stop all plugins. None to stop Stop all network services. What is the command to do that? Unmount the user share file system. What is the command to do that? Unmount all the data disks. What is the command to do that? Stop the md-driver. What is the command to do that? Reboot. Or stop any I/O to the server, reboot and cancel the subsequent parity check. Luckily I do know the shutdown or restart command, and thats all i could do and cancel the parity check upon restart. If I am not mistaken there should be a 'sync' somewhere in the steps above (unless its apart of one of those comments above). But would be nice to know how to gracefully do this via command line, or maybe you could supply a script we could keep in the root of the flash drive? Quite a few of the options in the WebGui can be executed from the commandline, but not a graceful shutdown/reboot as the WebGui would (is it top secret?) Loading a VM with basic version running Beta12a (I wanted to go pretty far back), clicked apply to every setting in the webgui. Loaded another VM with basic version running RC16c, clicked apply to every setting in the webgui. Results of differences: ident.cfg New value(s) in RC16c SYS_MODEL="" SYS_SLOTS="" No idea what they are used for. One would only see these new entries if you clicked 'Apply' in "Identification" section disk.cfg old value removed from Beta12c startState="" and replaced with in RC16c startArray="no" Understood. network.cfg New value(s) in RC16c BONDING="no" BONDING_MODE="1" Understood. One would only see these new entries if you clicked 'Apply' in "Network Setting" section share.cfg New value(s) in RC16c fuse_remember="330" Not understood, as I though its the NFS Settings that receive this new parameter in the 'NFS' section in the WebGui (unless NFS/AFP/SMB setting all get written to share.cfg, then it would be understood, seems like they are, looking at it now.) One would only see this new entries if you clicked 'Apply' in the respective section without having to make any change to a setting. Not sure how to say this, expect unwelcome surprise to not have these entries added immediately upon an upgrade (and booting up) in the version they stem from. syslogsmbsettingsclickedapply.txt
July 17, 201312 yr Hi all, A small issue I've just noticed - the system won't power off on shutdown - I'm running rc16b now but I did not see anything new in rc16c release notes that may change this behaviour. It used to work with earlier releases - at least with those I've tried (I had rc12a before) MB is an older Intel DP35DP / E2160 / 5x1.5T HDD + 3 TB parity.
July 17, 201312 yr Hi all, A small issue I've just noticed - the system won't power off on shutdown - I'm running rc16b now but I did not see anything new in rc16c release notes that may change this behaviour. It used to work with earlier releases - at least with those I've tried (I had rc12a before) MB is an older Intel DP35DP / E2160 / 5x1.5T HDD + 3 TB parity. update to rc16c and let us know if it still happens
July 18, 201312 yr so quiet around here. Are there no more problems. Could we really be THAT close to final? Are we just waiting on documentation?!?!?!
July 18, 201312 yr Hope so based on what i can see all issues reported are plug ins or add ons. So it seems that we might be waiting for some final confermation? Sent from my YP-G70 using Tapatalk 2
July 18, 201312 yr I'll say that rc16c seemed to resolve my Transport Endpoint Not Connected issues with Plex. I was able to recently add my entire music library (approx 9,000 tracks) and scan it on Plex with nary an issue, when previously even scanning a few TV shows at a time (maybe 200 episodes?) would cause the issue to happen.
July 18, 201312 yr Hi all, A small issue I've just noticed - the system won't power off on shutdown - I'm running rc16b now but I did not see anything new in rc16c release notes that may change this behaviour. It used to work with earlier releases - at least with those I've tried (I had rc12a before) MB is an older Intel DP35DP / E2160 / 5x1.5T HDD + 3 TB parity. update to rc16c and let us know if it still happens Performed the update to RC16C - no problems noticed so far, but it has the same behaviour .... on shutdown the system doesn't power-off. Could be some ACPI / APIC related (had no time to check further). Certainly, it did work up to 12/12a and with all previous betas I've been using during the last couple of years (far as I remember). Forgot to mention earlier - stock unRaid/no addons.
July 18, 201312 yr my system used to basically reboot, not shutdown with previous versions of unraid. with 16c it stays shutdown now on mine?
July 18, 201312 yr Performed the update to RC16C - no problems noticed so far, but it has the same behaviour .... on shutdown the system doesn't power-off. Could be some ACPI / APIC related (had no time to check further). Certainly, it did work up to 12/12a and with all previous betas I've been using during the last couple of years (far as I remember). Forgot to mention earlier - stock unRaid/no addons. Some people also have problems with ACPI stuff since RC12a, see here: http://lime-technology.com/forum/index.php?topic=28055.msg248671#msg248671 I know sleep is not supported but still... It seems that something related has changed from 12a on :-(
July 19, 201312 yr I know sleep is not supported but still... It seems that something related has changed from 12a on :-( S3 sleep in the past was problematic because of "buggy" motherboard bios. I didn't want to get into a situation of having to write all kinds of specialized modules to support every motherboard out there. These days the situation is much better and I agree that S3 sleep support should be "standard". Should be...
July 19, 201312 yr One fairly minor point... running rc15a I used to rely on the PC 'beep' when unRAID came up, but now I've had to add it as a command in the go script. Can this be reinstated? Not yet upgraded to 16c, so if anybody can verify this is the case therein, I will upgrade soon enough anyway. Thanks. m
July 19, 201312 yr ... Not yet upgraded to 16c, so if anybody can verify this is the case therein, I will upgrade soon enough anyway. Thanks. ... As a gerneral rule of thumb you shouldnt be running an rc unless its the latest and people arent going to pay a lot of attention if your not. Join the party... upgrade
July 19, 201312 yr OK - I'm on rc16c. It still does not beep on boot as 4.7 used to, which is very useful in headless operation in determining when the box is "up". I've manually added beep;beep;beep at the end of my go script for now. m
July 20, 201312 yr OK - I'm on rc16c. It still does not beep on boot as 4.7 used to, which is very useful in headless operation in determining when the box is "up". Having been running V5 since beta4, I no longer remember what 4.x used to do. However, I think that the definition "up" is a little uncertain. Is it when disk shares are available, user shares available, the emhttp interface is available, or all "startup activity" has completed? Termination of the 'go' script does not, necessarily, coincide with any of these. I know that, in my system, some of my addons are active long before the emhttp interface starts responding. I can access unMENU before emhttp. I believe that this occurs because of the way that the events interface works, and the fact that I run a number of addons. I've manually added beep;beep;beep at the end of my go script for now. This may well be the best solution. Add the beeps where they match your definition of "up". This could be in the 'go' script, or it could be on one of the events.
July 20, 201312 yr Good points all PeterB. I shall do some more digging, and probably best left to each user's requirements.
July 20, 201312 yr It still does not beep on boot as 4.7 used to, which is very useful in headless operation in determining when the box is "up". The beep isn't actually the system coming up but is the network daemon detecting a change (e.g. connecting). I always found I had to wait for a bit until emhttp & unmenu were actually started. I've confirmed this by unplugging & reconnecting the ethernet cable and it beeps on both events.
July 20, 201312 yr cfmjohn, I think the thing that is missing that it used to do was an initial beep when the network was first established. That is the beep that it no longer does. Good to know the network down/up events still beep.
July 20, 201312 yr Is there a fix for the user share "free" size being incorrect? Can't seem to find an answer on here. Apart from that above, the RC16c has been working fine for me. Thanks
July 22, 201312 yr Is there a fix for the user share "free" size being incorrect? Can't seem to find an answer on here. Apart from that above, the RC16c has been working fine for me. Thanks I thought Tom said he fixed that.
July 22, 201312 yr Should the Cache Drive also conform to these instructions? Click on each disk link on the Main page and examine the Partition format field. If you see "MBR: error", or "MBR: unknown" for any disk, do not Start the array; instead post your finding in the Forum announcement thread for this release. If everything looks ok, click Start to bring the array on-line. After upgrading from 5rc8a to 5rc16c, my cache drive says: Partition format: unknown
July 23, 201312 yr I thought Tom said he fixed that. I guess it's not totally fixed. 1 share is still displaying incorrect, despite being restricted to 1 drive. Manually running the mover subtracts the cache size from the total, but it's still including free space available of all drives. When I posted the above it was happening to a few of my shares. Now it's isolated to just the 1. I'll keep an eye on it while I wait for the next rc/final release. Thanks.
July 23, 201312 yr Should the Cache Drive also conform to these instructions?Yes... it should After upgrading from 5rc8a to 5rc16c, my cache drive says: Partition format: unknown Seek help from Tom at lime-tech. Joe L.
Archived
This topic is now archived and is closed to further replies.