On 9/2/14 16:15 , David Finster via smartos-discuss wrote: > Hi All > > On Monday I reverted the SmartOS distro being used back to vanilla Joyent and > upgraded to the latest version. > > I was using the UQ EAIT builds found here: http://imgapi.uqcloud.net/builds > > So far the issue hasn’t reoccured but I’ll keep an eye on it. arekinath > (maintainer of the builds) thought it might be related to the AMD patches > that he introduced, but it appears to be resolved all the same. Will keep an > eye on it though.
If you end up finding something like this again, an additional thing to end up checking out and looking at are the in-kernel per-CPU ring buffers. They'll be able to tell us a lot of information about what's happened. I don't recall off hand whether or not they include the block information, but should help to figure out what's going on. In general, the use of pstack/mdb on the user land process will probably cause the threads in question to basically exit the kernel and re-enter so that's why running those will sometimes kick things off again. Robert ------------------------------------------- smartos-discuss Archives: https://www.listbox.com/member/archive/184463/=now RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00 Modify Your Subscription: https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb Powered by Listbox: http://www.listbox.com
