hi: we have a freeradius 3.0.9 server for testing with very low traffic. today I found it was dead. the last log lines like below: Mon Oct 12 08:56:55 2015 : Warning: rlm_sql (sql): You probably need to lower "min" Mon Oct 12 08:56:55 2015 : Info: rlm_sql (sql): Closing connection (1440): Hit idle_timeout, was idle for 286 seconds Mon Oct 12 08:56:55 2015 : Info: rlm_sql (sql): Closing connection (1439): Hit idle_timeout, was idle for 285 seconds Mon Oct 12 08:56:55 2015 : Info: rlm_sql (sql): Opening additional connection (1441), 1 of 32 pending slots used Mon Oct 12 08:56:55 2015 : Info: rlm_sql (sql): 0 of 1 connections in use. Need more spares Mon Oct 12 08:56:55 2015 : Info: rlm_sql (sql): Opening additional connection (1442), 1 of 31 pending slots used Mon Oct 12 08:56:58 2015 : Info: rlm_sql (sql): 0 of 2 connections in use. Need more spares Mon Oct 12 08:56:58 2015 : Info: rlm_sql (sql): Opening additional connection (1443), 1 of 30 pending slots used Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): Closing connection (1441): Hit idle_timeout, was idle for 71 seconds Mon Oct 12 08:58:35 2015 : Warning: rlm_sql (sql): You probably need to lower "min" Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): Closing connection (1443): Hit idle_timeout, was idle for 64 seconds Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): Closing connection (1442): Hit idle_timeout, was idle for 63 seconds Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): Opening additional connection (1444), 1 of 32 pending slots used Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): 0 of 1 connections in use. Need more spares Mon Oct 12 08:58:35 2015 : Info: rlm_sql (sql): Opening additional connection (1445), 1 of 31 pending slots used Mon Oct 12 08:58:42 2015 : Error: (5838) Ignoring duplicate packet from client wireless-ap port 1482 - ID: 66 due to unfinished request in component authenticate module eap_peap the radiusd service die immediately after writing last log. checking mysql log I found before dead it successfully authenticate a user at 08:57:31. search the mail list I found the issue "unavailable database makes freeradius 3.0.9 server unresponsive" seems related. does the bug fixed in 3.0.10 or I need to log more radiusd info to track the issue? thanks a lot for help!! Regards, tbskyd
Hi,
does the bug fixed in 3.0.10 or I need to log more radiusd info to track the issue?
3.0.10 was released because lofs of bugs were fixed. the first thing to do is try the latest version to see if that bug has been already squashed. if not, THEN do the debugging/tracing etc alan
2015-10-22 16:27 GMT+08:00 <A.L.M.Buxey@lboro.ac.uk>:
Hi,
does the bug fixed in 3.0.10 or I need to log more radiusd info to track the issue?
3.0.10 was released because lofs of bugs were fixed. the first thing to do is try the latest version to see if that bug has been already squashed. if not, THEN do the debugging/tracing etc
alan -
ok I will try 3.0.10. any suggestion about radiusd debug setting so I can get useful info if the situation happen again? the server crashed after online about a month, so I think it is not easy to trigger in my environment. thanks for the hint. Regards, tbskyd
On Thu, Oct 22, 2015 at 4:19 PM, d tbsky <tbskyd@gmail.com> wrote:
2015-10-22 16:27 GMT+08:00 <A.L.M.Buxey@lboro.ac.uk>:
Hi,
does the bug fixed in 3.0.10 or I need to log more radiusd info to track the issue?
3.0.10 was released because lofs of bugs were fixed. the first thing to do is try the latest version to see if that bug has been already squashed. if not, THEN do the debugging/tracing etc
alan -
ok I will try 3.0.10. any suggestion about radiusd debug setting so I can get useful info if the situation happen again? the server crashed after online about a month, so I think it is not easy to trigger in my environment.
thanks for the hint.
Assuming you compile your own binary, OR your package comes with debug packages, see "panic_action" on radiusd.conf -- Fajar
participants (3)
-
A.L.M.Buxey@lboro.ac.uk -
d tbsky -
Fajar A. Nugraha