You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When running under -mode psicofee after installing tcoffee and related stuff, parrot crashes out to the singularity image with signal 11.
The details of where it is dying are vague, as tcoffee generates perl (and possibly other stuff) dynamically and runs it, so I can't tell where things are dying, but it's happening while it tries to get "templates" for sequences. This is a non-deterministic process, it appears, so sometimes it dies on the first template search and sometimes it does a few before dying. I can't tell where it's dying, only that it is fatal for parrot. I regret that tracing things to find out exactly what happened proved more difficult, but if I have time I might try a better "minimum" case to duplicate this.
On the bright side, I think its using the cvmfs stuff just fine, and it may just be a matter of how it is responding to error messages from unsuccessful perl scripts poorly, but it isn't related to #8 (applying my fix for that doesn't fix stuff).
The text was updated successfully, but these errors were encountered:
Singularity does support cvmfs but with fuse3 in the 3.4 release. I think there is some restriction on the host or the kernel version. I will need to look at that.
When running under
-mode psicofee
after installing tcoffee and related stuff, parrot crashes out to the singularity image with signal 11.The details of where it is dying are vague, as tcoffee generates perl (and possibly other stuff) dynamically and runs it, so I can't tell where things are dying, but it's happening while it tries to get "templates" for sequences. This is a non-deterministic process, it appears, so sometimes it dies on the first template search and sometimes it does a few before dying. I can't tell where it's dying, only that it is fatal for parrot. I regret that tracing things to find out exactly what happened proved more difficult, but if I have time I might try a better "minimum" case to duplicate this.
On the bright side, I think its using the cvmfs stuff just fine, and it may just be a matter of how it is responding to error messages from unsuccessful perl scripts poorly, but it isn't related to #8 (applying my fix for that doesn't fix stuff).
The text was updated successfully, but these errors were encountered: