Using:
FreeBSD 4.9 RELEASE,
postgresql-7.3.4_1,
dbmail-postgresql-1.2.1,
exim-4.24_2
dbmail-adduser talks to the DB, Exim+dbmail-smtp get messages into the DB.
Attempting to invoke dbmail-imapd or dbmail-pop3d result in runaway,
wherein each child fails to connect to the DB, exits fatal, is replaced
by another child.
Tried so far, with no joy:
1) bumped pgsql to -B 512 -N 250 to make more buffers and connections
available (default was 32).
2) Backed down the number of 'childs' in dbmail.conf to *one*.
Have looked at pop3d.c and imapd.c, suspect that even though built for
postgreSQL, there may be a vestige of MySQL'ism, ELSE *something* not as
pgsql-aligned as dbmail-smtp and dbmail-adduser (both of which work as
expected). .....But when I last programmed in C it was being compiled
under CP/M on a 4 MHz Z-80 ..... so ...
Any chance this relates to these daemons at one time using 100% CPU load
(reported last spring IIRC, but apparently fixed - at least w/r the more
common MySQL versions).
Bill Hacker
FreeBSD 4.9 RELEASE,
postgresql-7.3.4_1,
dbmail-postgresql-1.2.1,
exim-4.24_2
dbmail-adduser talks to the DB, Exim+dbmail-smtp get messages into the DB.
Attempting to invoke dbmail-imapd or dbmail-pop3d result in runaway,
wherein each child fails to connect to the DB, exits fatal, is replaced
by another child.
Tried so far, with no joy:
1) bumped pgsql to -B 512 -N 250 to make more buffers and connections
available (default was 32).
2) Backed down the number of 'childs' in dbmail.conf to *one*.
Have looked at pop3d.c and imapd.c, suspect that even though built for
postgreSQL, there may be a vestige of MySQL'ism, ELSE *something* not as
pgsql-aligned as dbmail-smtp and dbmail-adduser (both of which work as
expected). .....But when I last programmed in C it was being compiled
under CP/M on a 4 MHz Z-80 ..... so ...
Any chance this relates to these daemons at one time using 100% CPU load
(reported last spring IIRC, but apparently fixed - at least w/r the more
common MySQL versions).
Bill Hacker