Upon further review, it appears the problem lies in how the genome browser
server is handling the https request. We moved the example file to a
non-secure server and discovered that toggling the URL between http and
https toggles the connection error problem.

Was there some change late Monday that would affect the https protocol? As I
said, the https URL was working fine up until Monday.

Thanks,
Curt


On 9/24/09 10:52 AM, "Curtis Jamison" <[email protected]> wrote:

> Hi All -
> 
> As of Tuesday, I am suddenly encountering an issue with my bigBed tracks. The
> track info uploads fine, but when I try to go to the browser I get the
> following error:
> 
>     Couldn't set connection database to hg19
>     MySQL server has gone away
> 
> I'm getting this from tracks I know are fine (e.g. they were working Monday).
> I also made a new track with the example data.
> 
> My apache access log shows the genome browser is trying to access the file:
> 
> 205.142.199.52 - - [24/Sep/2009:10:22:52 -0400] "GET /cchmctracks/myBigBed.bb
> HTTP/1.1" 206 8192 "-" "genome.ucsc.edu/net.c"
> 205.142.199.53 - - [24/Sep/2009:10:26:40 -0400] "HEAD /cchmctracks/myBigBed.bb
> HTTP/1.1" 200 - "-" "genome.ucsc.edu/net.c"
> 205.142.199.53 - - [24/Sep/2009:10:27:27 -0400] "HEAD /cchmctracks/myBigBed.bb
> HTTP/1.1" 200 - "-" "genome.ucsc.edu/net.c"
> 205.142.199.53 - - [24/Sep/2009:10:28:53 -0400] "HEAD /cchmctracks/myBigBed.bb
> HTTP/1.1" 200 - "-" "genome.ucsc.edu/net.c"
> 
> Interestingly, the error message appears before the request appears in the
> tail -f listing of the access log.
> 
> Has anyone ever seen this before? At this point I suspect the IT people did
> something Monday night to increase the latency of the network proxy, but would
> love to hear other ideas.
> 
> Thanks,
> Curt
> 
> 
> 


_______________________________________________
Genome maillist  -  [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to