limetech Posted June 7, 2013 Share Posted June 7, 2013 I wake up in the middle of the night and wife says, "You were having a nightmare, yelling something about 'stale handles', wtf does that mean?" Link to comment
AndroidCat Posted June 7, 2013 Share Posted June 7, 2013 Tom, is that problem related to stale file handle? I am stuck on rc10 due to this. Apparently it's also broken in rc13: http://lime-technology.com/forum/index.php?topic=25739.0 "Some media players and receivers can see NFS share as empty" Link to comment
PeterB Posted June 8, 2013 Share Posted June 8, 2013 You were having a nightmare, yelling something about 'stale handles' Lol! Link to comment
unevent Posted June 8, 2013 Share Posted June 8, 2013 low level api vs. high level api? Link to comment
johnodon Posted June 10, 2013 Share Posted June 10, 2013 Specific to RC13... Has it been determined that the issue of clients (namely OpenELEC/XBMC) not being able to access NFS shares is related to "stale file handles"? if not can a separate issue be added to the list for that? I can provide any logs/testing required as I can reproduce the issue at will. All previous RCs/betas do not have this issue. John Link to comment
drawz Posted June 11, 2013 Share Posted June 11, 2013 There was an OpenELEC bug connecting to certain versions of NFS that I think was fixed in the latest release (within the last few days I believe). I posted links in another thread. Link to comment
johnodon Posted June 11, 2013 Share Posted June 11, 2013 There was an OpenELEC bug connecting to certain versions of NFS that I think was fixed in the latest release (within the last few days I believe). I posted links in another thread. You sir are absolutely correct!!! I just updated OpenELEC from v3.03 to v3.05 and also updated unRAID from RC12a to RC13. NFS shares are working as expected. W00t!!! I'll also post this in the RC13 thread. Thank you so much for the heads-up. John Link to comment
drawz Posted June 11, 2013 Share Posted June 11, 2013 Thank you so much for the heads-up. No prob. Enjoy! Link to comment
PeterB Posted June 13, 2013 Share Posted June 13, 2013 Just a quick post to confirm that my reproducible stale file handle error, resulting from running mkvmerge against user shares, does, indeed, appear to be resolved in rc14. Link to comment
limetech Posted June 13, 2013 Author Share Posted June 13, 2013 Just a quick post to confirm that my reproducible stale file handle error, resulting from running mkvmerge against user shares, does, indeed, appear to be resolved in rc14. Can I get a "Hallelujah!"? Link to comment
johnodon Posted June 13, 2013 Share Posted June 13, 2013 Just a quick post to confirm that my reproducible stale file handle error, resulting from running mkvmerge against user shares, does, indeed, appear to be resolved in rc14. Can I get a "Hallelujah!"? As long as another kernel is not introduced! Link to comment
AndroidCat Posted June 13, 2013 Share Posted June 13, 2013 Tom, I am still tirelessly asking you if "my issue" (see one of the first posts in this thread) has ever been identified and worked on? Sorry, I'm unable to test it this week myself. Link to comment
Dephcon Posted August 21, 2013 Share Posted August 21, 2013 Sorry to bump this back from page two, but I encountered a stale file handle issue on my tv share when i forced sickbeard to do a full re-scan. Here's what I had in fstab on my ubuntu 13.04 box 10.81.0.13:/mnt/user/TV\040Shows /mnt/unraid/tv nfs rw,nosuid,rsize=32768,wsize=32768,timeo=300,nfsvers=3 0 0 I added "lookupcache=none,noac" to the options and it seemed to fix it, but the performance went to shit (20MB/s write down from 80MB/s). Link to comment
limetech Posted August 21, 2013 Author Share Posted August 21, 2013 Sorry to bump this back from page two, but I encountered a stale file handle issue on my tv share when i forced sickbeard to do a full re-scan. Here's what I had in fstab on my ubuntu 13.04 box 10.81.0.13:/mnt/user/TV\040Shows /mnt/unraid/tv nfs rw,nosuid,rsize=32768,wsize=32768,timeo=300,nfsvers=3 0 0 I added "lookupcache=none,noac" to the options and it seemed to fix it, but the performance went to shit (20MB/s write down from 80MB/s). What do you have 'fuse_remember' set to? Link to comment
Dephcon Posted August 21, 2013 Share Posted August 21, 2013 if that's an unraid tune-able, then I haven't changed it from the default. If you need more info please let me know and I can work with you on it. I'm determined to migrate sab/sick/couch/etc off unraid into a dedicated linux system and really want NFS to work for me. Link to comment
Recommended Posts