sendmail 8.11.1 on 11.0 or 11.11

[Scott Donaldson]

Hi,

I needed to establish recipient whitelisting so I used the delay_checks feature, added some entries to the access map and implemented. My whitelisting works...but now sendmail 'throws' the following errors up to 4 times a day (which it did not do before I rebuild my sendmail.cf file with the delay_checks feature). I've searched for solutions and it seems like it may simply be a transient error that is a side effect of the various implementations of the accept() system call.

Anyone see this error and find a solution? Or decide to live with it?

The full error text is below:


Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: SYSERR(root): getrequests: accept: No buffer space available
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 0: fl=0x0, mode=20666: CHR: size=0
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 1: fl=0x1, mode=20666: CHR: size=0
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 2: fl=0x1, mode=20666: CHR: size=0
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 3: fl=0x2, mode=140000: SOCK [[UNIX: /var/spool/sockets/pwgr/client13858]]->(Socket is not connected)
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 4: fl=0x10001, mode=10666: FIFO: size=4096
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 5: fl=0x2, mode=140000: SOCK [0.0.0.0]/25->(Socket is not connected)
Jul 3 07:38:05 myhost sendmail[13858]: NOQUEUE: 6: fl=0x2, mode=140000: SOCK [0.0.0.0]/587->(Socket is not connected)
Jul 3 07:38:05 myhost sendmail[13858]: accepting connections again for daemon MTA

Scott.