Re: [OMPI devel] orted --seed and orte_init()

On Feb 3, 2007, at 10:35 AM, Ralph Castain wrote: Something did occur to me that *might* help with the problem of detecting when the seed is running. There is an option to orted "-- report- uri pipe" that will cause the orted to write it's uri to the specified pipe. This comes after the

Re: [OMPI devel] orted --seed and orte_init()

On Feb 3, 2007, at 6:51 AM, Ralph Castain wrote: On 2/2/07 8:44 AM, "Greg Watson" wrote: We're launching a seed daemon so that we can get registry persistence across multiple job launches. However, there is a race condition between launching the daemon and the first

Re: [OMPI devel] orted --seed and orte_init()

Something did occur to me that *might* help with the problem of detecting when the seed is running. There is an option to orted "-- report-uri pipe" that will cause the orted to write it's uri to the specified pipe. This comes after the orted has completed orte_init, and so it *should* be ready at

Re: [OMPI devel] orted --seed and orte_init()

On 2/2/07 8:44 AM, "Greg Watson" wrote: > We're launching a seed daemon so that we can get registry persistence > across multiple job launches. However, there is a race condition > between launching the daemon and the first call to orte_init() that > can result in a bus