actually it is possible to disable transaction? i believe the cluster will handle auto-recommit except for transactional commit.
On Saturday, July 20, 2013 4:39:01 PM UTC+7, vince wrote: > > i think the framework errors are from web2py session update into mysql, > we've already use session.forget to minimize the problem but still we need > to update session sometimes. > > we have already catching all update request and do retry ourself. since > mariadb galera cluster is synchronous replication i guess it's no way to > avoid deadlock during update. > > is there any solution to do custom update retry on web2py session? it > would be ideal if dal have an option to auto retry update query. > > > > On Friday, July 19, 2013 7:58:30 PM UTC+7, Alan Etkin wrote: >> >> i do not know which function actually causing this problem. the >>> application we testing are mostly select query so most likely it's from the >>> web2py session update query. >> >> >> I think this is a workflow issue. Perhaps your db error occurs after >> executing the action logic. In that case, this pseudocode might help: >> >> At a web2py controller: >> >> try some DAL db i/o command >> force a commit (with db.commit()) >> if there's a driver exception wait for a few seconds and retry >> if the second try fails, return a message >> else return the query output >> >> BTW: How did you implement my previous pseudocode? What was the error >> output? >> > -- --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
