Fatal Trap 12 using HAST on shutdown

[hblandford]

Hi All,

I have HAST running with ucarp (on a lagg) to swap a ZFS pool between primary and secondary nodes.

I can stop and start ucarp and watch the role swap happily from one box to the other time after time.

My problem comes when I shutdown the master. I run # shutdown -r now and away it goes. On the secondary, it will detect it needs to become the master. However, about 33% or 50% of the time it will die with a Fatal Trap 12 :-(

It doesn't matter which way round I do it. Both boxes suffer the same problem. I am running RELENG_8 from 2 days ago. I have only be playing with this since 8.1R but it has happened on all versions I have tried.

Anyone have any suggestions on what I should be looking at?

Thanks,

Hugh

 

[phoenix]

Showing the trap message would be helpful, as it will show the process(es) running when the trap happens.

As a test, have you tried it using devd and carp(4) instead of ucarp? Michael Lucas did a couple of blog posts on it, using a variation of the scripts/configs I came up with for this when HAST was initially announced.