Обсуждение: Re: [HACKERS] please?
> Bruce Momjian wrote: > > Sharing file systems. Good point. You could have a table you use to > > lock. Lock the table, view the value, possibly modify, and unlock. > > This does not handle the case where someone died and did not remove > > their entry from the lock table. > > You can always write the modification time to the table as well and if > it's "too old", then try to override it. > Assuming you can set a reasonable "too old" time. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
> > > Bruce Momjian wrote: > > > Sharing file systems. Good point. You could have a table you use to > > > lock. Lock the table, view the value, possibly modify, and unlock. > > > This does not handle the case where someone died and did not remove > > > their entry from the lock table. > > > > You can always write the modification time to the table as well and if > > it's "too old", then try to override it. > > > > Assuming you can set a reasonable "too old" time. > There may be many partial workarounds, depending on the application, but there seems to be no robust way to have a failed lock right now. Perhaps in a future version will PQrequestCancel be able to terminate a waiting-for-lock state? Pablo
Pablo Funes <pablo@cs.brandeis.edu> writes: > Perhaps in a future version will PQrequestCancel be able to terminate > a waiting-for-lock state? Seems like a reasonable suggestion. It's too late to consider this for 6.5 (we were supposed to freeze the feature list quite a while back) but I support putting it on the TODO list for a future release. regards, tom lane
> Pablo Funes <pablo@cs.brandeis.edu> writes: > > Perhaps in a future version will PQrequestCancel be able to terminate > > a waiting-for-lock state? > > Seems like a reasonable suggestion. It's too late to consider this for > 6.5 (we were supposed to freeze the feature list quite a while back) > but I support putting it on the TODO list for a future release. > > regards, tom lane > Added: * Allow PQrequestCancel() to terminate when in waiting-for-lock state -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
Cool! :-) > > > Pablo Funes <pablo@cs.brandeis.edu> writes: > > > Perhaps in a future version will PQrequestCancel be able to terminate > > > a waiting-for-lock state? > > > > Seems like a reasonable suggestion. It's too late to consider this for > > 6.5 (we were supposed to freeze the feature list quite a while back) > > but I support putting it on the TODO list for a future release. > > > > regards, tom lane > > > > Added: > > * Allow PQrequestCancel() to terminate when in waiting-for-lock state >
Added to TODO: * PQrequestCancel() be able to terminate backend waiting for lock > Pablo Funes <pablo@cs.brandeis.edu> writes: > > Perhaps in a future version will PQrequestCancel be able to terminate > > a waiting-for-lock state? > > Seems like a reasonable suggestion. It's too late to consider this for > 6.5 (we were supposed to freeze the feature list quite a while back) > but I support putting it on the TODO list for a future release. > > regards, tom lane > -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026