Hi. In one word: no James doesn't support the failover mechanism as it is, at the moment. We have the same problem and we are going to use the following configuration: - 2 James instances with a loadbalancer. - A pop3 lock patch we developed (for a single james) - Separated spool and outgoing repositories - Common inbox repository.
If an instance fails (i.e. telnet on 25 and on 110 dies), a daemon copies the spool from an instance to the other and it should keep going. Hope it helps Cheers Marcello > -----Messaggio originale----- > Da: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] > Inviato: martedì 28 marzo 2006 3.10 > A: [email protected] > Oggetto: HA deployment of James > > Hi, > > We currently have James running in a production environment > (on a linux blade). We are building out a HA environment by > adding another blade that shares the same NAS head and hence > has the NFS mount point. > > Is there a way to run James in a hold-cold manner so that if > the primary fails the secondary can take over? Is there a way > to run James in a clustered manner? Is the locking mechanism > to read the emails from the filesystem based on in-memory > locking or is there support for filesystem based locking? In > other words is it possible for 2 james JVMs to go after the > same set of records on the filesystem? > > Thanks for your time. > > Rgds, > -Mouli > > > IB Arch: CBB NY > > > > ,: [EMAIL PROTECTED] > (: +1 212-622-3734 > GDP: 622-3734 > È: +1 201-637-3603 > > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
