[asterisk-bugs] [JIRA] (ASTERISK-26720) Skylake Compatibility

Adam Chandler (JIRA) noreply at issues.asterisk.org
Mon Jan 16 08:56:10 CST 2017


Adam Chandler created ASTERISK-26720:
----------------------------------------

             Summary: Skylake Compatibility 
                 Key: ASTERISK-26720
                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-26720
             Project: Asterisk
          Issue Type: Bug
      Security Level: None
         Environment: TE820BF and Skylake
            Reporter: Adam Chandler


>From customer:
"We tried the TE820BF boards in 3 identical servers with X11SSH-LN4F boards ( 
https://www.supermicro.com/products/motherboard/Xeon/C236_ 
C232/X11SSH-LN4F.cfm), with latest 2.0 BIOS. All three servers/boards pairs exhibit the same behaviour : poor dahdi_test results, lots of patlooptest errors, no issue when trying with a TE420B instead.
So I don't believe any current TE820B would fare any better.


Note that your engineering team might get in touch with Bryan Z. from 
Supermicro support (support at supermicro.com) who's in turn in touch with 
their own engineering team regarding this problem. "

Further investigation seems to show an issue with dahdi_test : root at tchad4:~ dahdi_test 
Opened pseudo dahdi interface, measuring accuracy... 

32.457% 36.981% 14.045% 44.101% 39.238% 35.184% 28.854% 36.959% 
60.502% 44.674% 40.474% 34.702% 37.622% 33.670% 34.920% 50.408% 
42.547% 21.976% 64.419% 42.443% 42.241% 61.834% 30.443% 63.014% 
28.280% 35.743% ^C 
--- Results after 26 passes --- 
Best: 64.419% -- Worst: 14.045% -- Average: 39.912695% 
Cummulative Accuracy (not per pass): 39.913


When in this setup the customer experiences HDLC aborts and audio distortion. When customer tests the TE820BF cards in much older servers, the issue is not present. 

My question is can we test this same setup for compatibility issues? Suggestions on where to take this next or what logs we can gather?



--
This message was sent by Atlassian JIRA
(v6.2#6252)



More information about the asterisk-bugs mailing list