Page 1 of 1

Xeoma 17.5.5 Linux 64 client won’t start.

PostPosted: Fri May 19, 2017 9:42 am
by jonbanjo
I tried the earlier 2016-12-26 version and was happy enough with the software to opt for the std version. The only problem (temporarily resolved by using the std older version which, luckily I kept) is that I can’t get 17.5.5 to start on my pcs.

I think I can narrow this down to the client portion but at present can’t offer much else in the way of feedback. If I try to start in a terminal, the result is a lock up that ctrl C will not break out of. There are no error messages or other feedback.

Systems tried are OpenSUSE Leap 42.1 and 42.2 both using the Plasma5 desktop.

Re: Xeoma 17.5.5 Linux 64 client won’t start.

PostPosted: Mon May 22, 2017 3:18 pm
by Admin_N
Thanks for having chosen Xeoma!

Sorry about the problem you've encountered. We'll check it now in our lab environment. Are there any specific modules that you use? IDK, visitors counter, or anything like that? That information would help a lot.
Anyway, we'll come back here when we have results.

P.S. By the way, "old" version is kept after the update in Xeoma's Oldversions folder.

Re: Xeoma 17.5.5 Linux 64 client won’t start.

PostPosted: Tue May 23, 2017 6:42 am
by jonbanjo
Thanks. I think I've found the problem. Starting first time as a normal user, strace shows this seemingly endless loop

Code: Select All Code
open("/tmp/FelenasoftFFmpegUnpackMutex2405161959", O_RDWR|O_CREAT|O_TRUNC, 0777) = 9
chmod("/tmp/FelenasoftFFmpegUnpackMutex2405161959", 0777) = 0
open("/var/lock/FelenasoftFFmpegUnpackMutex2405161959", O_RDWR|O_CREAT|O_TRUNC, 0777) = -1 EACCES (Permission denied)
open("/tmp/FelenasoftFFmpegUnpackMutex2405161959", O_RDWR) = 10
fcntl(10, F_SETLK, {type=F_UNLCK, whence=SEEK_CUR, start=0, len=100}) = 0
close(10)                               = 0
nanosleep({0, 500000000}, NULL)         = 0


If I start as root, the lock files (there seem to be 3) get written, I can then start as a normal user.

Re: Xeoma 17.5.5 Linux 64 client won’t start.

PostPosted: Tue May 23, 2017 2:18 pm
by Admin_P
Thank you very much for all the additional information. Our Testing Department is investigating the issue.