Hi, I've just upgraded one of my servers to 3.0.2o and have just noticed the following radius.log messages, although I'm fairly sure it's got nothing to do with the version release. My FR setup uses buffered-sql to offload sql writes via the details file. This config works fine on all my other servers ( puppet managed config) but on this one I've noticed 1). I see the following messages in the log Thu Nov 21 15:24:55 2019 : WARNING: (86473) WARNING: Module rlm_sql became unblocked\ Thu Nov 21 15:25:53 2019 : Error: Unresponsive child for request 87765, in component accounting module sql 2). the buffered-sql process copies contents of file detail into detail.work leaving the source to capture subsequent packets.oes An ls -lt shows -rw------- 1 freerad freerad 31359093 Nov 21 15:51 detail -rw------- 1 freerad freerad 101566485 Nov 21 15:51 detail-filebeat -rw------- 1 freerad freerad 63282 Nov 21 14:22 detail.work -rw------- 1 freerad freerad 211324742 Nov 21 06:24 detail-filebeat.1 So detail.work hasn't moved since 14:22 I'm guessing there's something in thre detail.work file causing the sql to fail. How can I debug the buffered-sql server to check this out without generating 3,000,000 logs for the rest of the server ? Rgds Alex
On Nov 21, 2019, at 10:56 AM, Alex Sharaz via Freeradius-Users <freeradius-users@lists.freeradius.org> wrote:
I've just upgraded one of my servers to 3.0.2o and have just noticed the following radius.log messages, although I'm fairly sure it's got nothing to do with the version release.
My FR setup uses buffered-sql to offload sql writes via the details file.
This config works fine on all my other servers ( puppet managed config) but on this one I've noticed
1). I see the following messages in the log
Thu Nov 21 15:24:55 2019 : WARNING: (86473) WARNING: Module rlm_sql became unblocked\
Which means that the SQL module is blocking the server.
2). the buffered-sql process copies contents of file detail into detail.work leaving the source to capture subsequent packets.oes
That's fine.
An ls -lt shows
-rw------- 1 freerad freerad 31359093 Nov 21 15:51 detail -rw------- 1 freerad freerad 101566485 Nov 21 15:51 detail-filebeat -rw------- 1 freerad freerad 63282 Nov 21 14:22 detail.work -rw------- 1 freerad freerad 211324742 Nov 21 06:24 detail-filebeat.1
So detail.work hasn't moved since 14:22
Because the SQL server is blocking, and isn't letting FreeRADIUS process requests.
I'm guessing there's something in thre detail.work file causing the sql to fail.
Probably not. The SQL server is *down*. If there was something in the packets that was bad, then the rlm_sql module would be
How can I debug the buffered-sql server to check this out without generating 3,000,000 logs for the rest of the server ?
You don't. You fix SQL so that it's responding to FreeRADIUS. Alan DeKok.
Hi When I've had this its because the detail file was not written to correctly when service killed or server shutdown less gracefully than it should have been. A quick session with vi and the detail.work file will find the culprit entry and then all will be well alan On Thu, 21 Nov 2019, 15:57 Alex Sharaz via Freeradius-Users, < freeradius-users@lists.freeradius.org> wrote:
Hi,
I've just upgraded one of my servers to 3.0.2o and have just noticed the following radius.log messages, although I'm fairly sure it's got nothing to do with the version release.
My FR setup uses buffered-sql to offload sql writes via the details file.
This config works fine on all my other servers ( puppet managed config) but on this one I've noticed
1). I see the following messages in the log
Thu Nov 21 15:24:55 2019 : WARNING: (86473) WARNING: Module rlm_sql became unblocked\
Thu Nov 21 15:25:53 2019 : Error: Unresponsive child for request 87765, in component accounting module sql
2). the buffered-sql process copies contents of file detail into detail.work leaving the source to capture subsequent packets.oes
An ls -lt shows
-rw------- 1 freerad freerad 31359093 Nov 21 15:51 detail -rw------- 1 freerad freerad 101566485 Nov 21 15:51 detail-filebeat -rw------- 1 freerad freerad 63282 Nov 21 14:22 detail.work -rw------- 1 freerad freerad 211324742 Nov 21 06:24 detail-filebeat.1
So detail.work hasn't moved since 14:22
I'm guessing there's something in thre detail.work file causing the sql to fail.
How can I debug the buffered-sql server to check this out without generating 3,000,000 logs for the rest of the server ?
Rgds
Alex - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
thanks Alan Rgds Alex Sent from my iPhone > On 22 Nov 2019, at 19:50, Alan Buxey <alan.buxey@gmail.com> wrote: > > Hi > > When I've had this its because the detail file was not written to correctly > when service killed or server shutdown less gracefully than it should have > been. > > A quick session with vi and the detail.work file will find the culprit > entry and then all will be well > > alan > >> On Thu, 21 Nov 2019, 15:57 Alex Sharaz via Freeradius-Users, < >> freeradius-users@lists.freeradius.org> wrote: >> >> Hi, >> >> I've just upgraded one of my servers to 3.0.2o and have just noticed the >> following radius.log messages, although I'm fairly sure it's got nothing to >> do with the version release. >> >> My FR setup uses buffered-sql to offload sql writes via the details file. >> >> This config works fine on all my other servers ( puppet managed config) but >> on this one I've noticed >> >> >> 1). I see the following messages in the log >> >> Thu Nov 21 15:24:55 2019 : WARNING: (86473) WARNING: Module rlm_sql became >> unblocked\ >> >> >> Thu Nov 21 15:25:53 2019 : Error: Unresponsive child for request 87765, in >> component accounting module sql >> >> >> 2). the buffered-sql process copies contents of file detail into >> detail.work leaving the source to capture subsequent packets.oes >> >> >> An ls -lt shows >> >> -rw------- 1 freerad freerad 31359093 Nov 21 15:51 detail >> -rw------- 1 freerad freerad 101566485 Nov 21 15:51 detail-filebeat >> -rw------- 1 freerad freerad 63282 Nov 21 14:22 detail.work >> -rw------- 1 freerad freerad 211324742 Nov 21 06:24 detail-filebeat.1 >> >> So detail.work hasn't moved since 14:22 >> >> I'm guessing there's something in thre detail.work file causing the sql to >> fail. >> >> >> How can I debug the buffered-sql server to check this out without >> generating 3,000,000 logs for the rest of the server ? >> >> >> Rgds >> >> Alex >> - >> List info/subscribe/unsubscribe? See >> http://www.freeradius.org/list/users.html > - > List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
participants (3)
-
Alan Buxey -
Alan DeKok -
Alex Sharaz