Hello,

Following a shutdown of our DNS server for maintenance purposes, we noticed the following messages in the online.log file (see below).

Our onconfig have this settings : NS_CACHE host=1800,service=900,user=900,group=900,sqlhosts=900

We are wondering what the best configuration strategy would be to prevent these messages from appearing in future. 

Thank you

Online.log :

21:44:20  WARNING: Detected slow or failing DNS service response 26 time(s).
05/16/2026 21:45:09   listener-thread@soc_be.c:3054: err = -25582: oserr = 0: errstr = from vmwbsapi22.wbs.**** to server ifx : Network connection is broken.

05/16/2026 21:45:12   listener-thread@soc_be.c:3054: err = -25582: oserr = 0: errstr = from vmwbsapi22.wbs.**** to server ifx : Network connection is broken.

05/16/2026 21:45:21   Checkpoint Completed:  duration was 0 seconds.

........

05/16/2026 22:04:37   An unexpected SQL error occurred during authentication (sql error=-213, isam
 error=-157). Check the sysuser database to see if it is corrupted.
05/16/2026 22:04:37   listener-thread: err = -951: oserr = 0: errstr = wss@vmwbsapi23.wbs.*****: Incorrect password or user wss@vmwbsapi23.wbs.******[vmwbsapi23] is not known on the database server.

05/16/2026 22:04:37   listener-thread@soc_be.c:3384: err = -25582: oserr = 0: errstr = : Network connection is broken.

....

As far as I know the HOST part of NS_CACHE serves both name-to-ip and ip-tp-name resolution purposes, so temporarily setting host=<really_high_value> probably should get you over such DNS maintenance window, provided that no client connects that hadn't connected yet before, i.e. whose name/ip combo isn't cached yet.

There might also be OS level DNS caching, but that would only even have an effect on Informix level if a NS_CACHE cache entry doesn't exist or expired.

HTH,
 Andreas