reggierat Posted January 9, 2015 Share Posted January 9, 2015 I have an ongoing issue with the S3 sleep Plugin. Occasionally the server does not seem to wake correctly from sleep, it appears 'hung' ie not accessible from the Web GUI, SMB, Putty etc plugging a monitor into the machine does not show any output either, But if i press the power button the machine will shutdown safely allowing me to power on again without triggering a parity check. The other issue i have is the server will sometimes not sleep when it should be expected to. This morning i got up to find it had been running all night. It appears that it is not loading all settings correctly on first boot. Looking at the syslog it is active because of SDi wihch is my cache drive. S3 sleep is set to exclude the Cache drive and i can see when the server booted up this evening that it did not include this setting. At 5.29am i togglen the exclude cache setting on and off and this fixed the issue and not i can see the server is going to sleep soon. syslog.zip Link to comment
Fireball3 Posted January 20, 2015 Share Posted January 20, 2015 S3 sleep is not always working. Means, there are setups where it works and others not. Search the forum for similar posts. I doubt it is the hard drive since it comes up OK when rebooting. Maybe your setup is one of those not working with S3? What version of S3 script do you use? Link to comment
bonienl Posted January 20, 2015 Share Posted January 20, 2015 Some people have reported that they need to run some specific commands before and/or after s3-sleep. When using the s3_sleep plugin of Dynamix it is possible to define these commands in the designated boxes. Which commands are exactly needed depends on your system, perhaps a forum search can help. Link to comment
reggierat Posted January 30, 2015 Author Share Posted January 30, 2015 Thanks for the replies, i realise the hard drives are a long shot, but this is really annoying me since on v5 this worked flawlessly. I did notice in the syslog it was complaining about an unclean shutdown and to run checkdisk on the flash drive. I ran checkdisk and then managed 10 days without issue. But the issue did re-occur after those 10 days. Again waking from sleep the system hung, no network shares, no telnet, no access when connecting monitor and keyboard. I have now performed a long format on my flash drive and set it up from scratch again and re-downloaded the latest dynamix plugins including s3 sleep. Apart from the above, i am still having the issue where on boot, S3 sleep is not setting the included/excluded disk parameters correctly eg upon installing the plugin and configuring it for the first time, i see the following in the syslog, drives are set correctly, it is including all disks and excluding cache and flash Jan 31 08:15:57 Tower s3_sleep: command-args=-C 1 -a -c -m 30 -D 0 Jan 31 08:15:57 Tower s3_sleep: action mode=sleep Jan 31 08:15:57 Tower s3_sleep: check disks status=yes Jan 31 08:15:57 Tower s3_sleep: check network activity=no Jan 31 08:15:57 Tower s3_sleep: check active devices=no Jan 31 08:15:57 Tower s3_sleep: check local login=no Jan 31 08:15:57 Tower s3_sleep: check remote login=no Jan 31 08:15:57 Tower s3_sleep: version=3.0.0 Jan 31 08:15:57 Tower s3_sleep: ---------------------------------------------- Jan 31 08:15:57 Tower s3_sleep: included disks=sdb sdc sdd sde sdf sdg sdh Jan 31 08:15:57 Tower s3_sleep: excluded disks=sda sdi Jan 31 08:15:57 Tower s3_sleep: ---------------------------------------------- Jan 31 08:15:57 Tower s3_sleep: s3_sleep process ID 12960 started, To terminate it, type: s3_sleep -q After shutting the system down and powering on again i see the following when the plugin is loaded, all disks are set to excluded Jan 31 08:19:17 Tower s3_sleep: command-args=-C 1 -a -c -m 30 -D 0 Jan 31 08:19:17 Tower s3_sleep: action mode=sleep Jan 31 08:19:17 Tower s3_sleep: check disks status=yes Jan 31 08:19:17 Tower s3_sleep: check network activity=no Jan 31 08:19:17 Tower s3_sleep: check active devices=no Jan 31 08:19:17 Tower s3_sleep: check local login=no Jan 31 08:19:17 Tower s3_sleep: check remote login=no Jan 31 08:19:17 Tower s3_sleep: version=3.0.0 Jan 31 08:19:17 Tower s3_sleep: ---------------------------------------------- Jan 31 08:19:17 Tower s3_sleep: included disks= Jan 31 08:19:17 Tower s3_sleep: excluded disks=sda sdb sdc sdd sde sdf sdg sdh sdi When i go into the plugins settings and toggle 'wait for array inactivity' from 'yes, exclude cache' to no, click apply and then back to 'yes, exlcude cache' i see the correct settings in the log Jan 31 08:21:37 Tower s3_sleep: command-args=-C 1 -a -c -m 30 -D 0 Jan 31 08:21:37 Tower s3_sleep: action mode=sleep Jan 31 08:21:37 Tower s3_sleep: check disks status=yes Jan 31 08:21:37 Tower s3_sleep: check network activity=no Jan 31 08:21:37 Tower s3_sleep: check active devices=no Jan 31 08:21:37 Tower s3_sleep: check local login=no Jan 31 08:21:37 Tower s3_sleep: check remote login=no Jan 31 08:21:37 Tower s3_sleep: version=3.0.0 Jan 31 08:21:37 Tower s3_sleep: ---------------------------------------------- Jan 31 08:21:37 Tower s3_sleep: included disks=sdb sdc sdd sde sdf sdg sdh Jan 31 08:21:37 Tower s3_sleep: excluded disks=sda sdi Jan 31 08:21:37 Tower s3_sleep: ------------------------- Link to comment
reggierat Posted February 4, 2015 Author Share Posted February 4, 2015 Managed to get a clear Syslog of the issue happening. I would say it is still maybe 1 in every 10-20 time it wakes from sleep that it hangs. Could someone please have a look at my syslog and offer some advice. I realise this is unsupported, but this did work flawlessly in v5 syslog.txt Link to comment
Fireball3 Posted February 4, 2015 Share Posted February 4, 2015 Can you collect a "good" boot also? Link to comment
reggierat Posted February 4, 2015 Author Share Posted February 4, 2015 Good Boot syslog.txt Link to comment
Fireball3 Posted February 5, 2015 Share Posted February 5, 2015 I'm no expert but what I can see in the "bad log" is this message repeating very often: "Tower kernel: r8169 0000:03:00.0 eth0: rtl_phyar_cond == 1 (loop: 20, delay: 25)." In the "good boot" the NIC is reporting just this: Feb 5 04:51:12 Tower kernel: r8169 0000:03:00.0 eth0: link down Feb 5 04:51:12 Tower kernel: r8169 0000:03:00.0 eth0: link up Can't say if this is the only reason but it is probably not bad to solve this first. Maybe a NIC restart/reset? (if there is something like that) ... Did some google-fu on that error and it seems it's an issue with the r8169 driver. Was there a driver update in v6? Thats what I could find. Maybe someone more knowledgable can chime in. Link to comment
reggierat Posted February 15, 2015 Author Share Posted February 15, 2015 Thank you for taking the time to look at this. I've done a little googling myself and it does appear to be Kernel related. I have also spoken to JonP who has stated we are moving to a new kernel next beta so fingers crossed. Still frustrating, just made it to 7 days before the issue re-occurred. If the updated Kernel doesn't improve things would you think something like http://www.mwave.com.au/product/intel-pro1000-gt-network-adapter-pci-pwla8391gtblk-aa38128 would be better? Link to comment
reggierat Posted February 20, 2015 Author Share Posted February 20, 2015 Issue still persists in latest Beta 13 Server lasted 4 days, put to sleep and woken roughly twice a day syslog attached syslog_2.txt Link to comment
reggierat Posted February 21, 2015 Author Share Posted February 21, 2015 Could anyone who uses s3_Sleep successfully with v6 post which NIC you are using. i am considering purchasing a new Intel NIC to see if it resolves my issue Link to comment
bonienl Posted February 22, 2015 Share Posted February 22, 2015 Could anyone who uses s3_Sleep successfully with v6 post which NIC you are using. i am considering purchasing a new Intel NIC to see if it resolves my issue s3_sleep works with me. Using Intel NIC. Would recommend to go the same way. Link to comment
Fireball3 Posted February 23, 2015 Share Posted February 23, 2015 While it is desireable to have a fully functional rig, you should answer yourself one question. What is the benefit from S3 over shutdown (S5)? In my case the gain is in the range of seconds. I decided to go with S5. Just some food for thought. Link to comment
reggierat Posted February 23, 2015 Author Share Posted February 23, 2015 While it is desireable to have a fully functional rig, you should answer yourself one question. What is the benefit from S3 over shutdown (S5)? In my case the gain is in the range of seconds. I decided to go with S5. Just some food for thought. You are right, most of the benefit is measured in seconds or the convenience of knowing if I click a song in the bedroom, the server 3 rooms away will wake, spin up and start playing in the bedroom. But since it was something I had working flawlessly in v5 I'm reluctant to take a step backwards so to speak. Today I have purchased an Intel 1000 pro gt for $45 so we'll see how we go Link to comment
Fireball3 Posted February 23, 2015 Share Posted February 23, 2015 I click a song in the bedroom, the server 3 rooms away will wake, spin up and start playing in the bedroom. Is this really the case? Who or what is sending the WoL packet to the server? Link to comment
reggierat Posted February 23, 2015 Author Share Posted February 23, 2015 The raspberry pi hidden behind the stereo in the bedroom. When I open yatse, change to bedroom and click music the server wakes up, clicking a track begins playback quickly afterwards. Link to comment
reggierat Posted March 2, 2015 Author Share Posted March 2, 2015 Going to make this one solved now. It has now been 8 days without a problem after purchasing http://www.scorptec.com.au/product/Networking_-_Wired/PCI_Cards/15392-PWLA8391GTBLK This is the longest i have gone since upgrading to Beta6 without having a crash waking from sleep. Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.