You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I recently had a device which exhausted my DHCPv4 pool. It took me a while to debug this, since there was nothing from odhcpd in the logs. The loglevel for odhcpd is set to 4 (WARN).
Of course odhcpd logs a lack of addresses - but a NOTICE level, so I didn't get it with my settings.
syslog(LOG_NOTICE, "Can't assign any IP address -> address space is full");
I think this is somewhat inconsistent with other messages at this level. In particular the following seem pretty normal and not "requiring special handling".
syslog(LOG_NOTICE, "Forward a RA on %s", c->name);
I believe "address space is full" is a fairly severe error condition (because address assignment fails and devices can't join the network) and should probably be logged at WARN or even ERROR. However, this messages was changed from WARN to NOTICE in d60f0a6 in response to #150.
RA and RS status messages seem less important to me and I suggest logging them at INFO. Note that this was the case for "Sending a RA" until it was promoted to NOTICE in 207f8e0 (probably for consistence with other messages). #191 even suggests DEBUG for "Sending a RA".
Implementing both changes I suggest would leave the NOTICE level pretty much unused and therefore useless. I'm fine with doing any one of them. I just want to be able to configure odhcpd to be silent unless something is broken - which currently is not quite possible. I think I set the loglevel to WARN because the RA messages I got every few minutes when I tried NOTICE.
@dedeckeh Since you made those changes to the log priorities back in 2019, maybe you can shed some light on the rationale and the best way forward?
I'll be happy to make a PR for the changes, if that's any help to you.
The text was updated successfully, but these errors were encountered:
I recently had a device which exhausted my DHCPv4 pool. It took me a while to debug this, since there was nothing from odhcpd in the logs. The
loglevel
for odhcpd is set to 4 (WARN).Of course odhcpd logs a lack of addresses - but a NOTICE level, so I didn't get it with my settings.
odhcpd/src/dhcpv4.c
Line 1098 in 5585b96
I think this is somewhat inconsistent with other messages at this level. In particular the following seem pretty normal and not "requiring special handling".
odhcpd/src/router.c
Line 889 in 5585b96
odhcpd/src/router.c
Line 957 in 5585b96
odhcpd/src/router.c
Line 987 in 5585b96
odhcpd/src/router.c
Line 1026 in 5585b96
I believe "address space is full" is a fairly severe error condition (because address assignment fails and devices can't join the network) and should probably be logged at WARN or even ERROR. However, this messages was changed from WARN to NOTICE in d60f0a6 in response to #150.
RA and RS status messages seem less important to me and I suggest logging them at INFO. Note that this was the case for "Sending a RA" until it was promoted to NOTICE in 207f8e0 (probably for consistence with other messages).
#191 even suggests DEBUG for "Sending a RA".
Implementing both changes I suggest would leave the NOTICE level pretty much unused and therefore useless. I'm fine with doing any one of them. I just want to be able to configure odhcpd to be silent unless something is broken - which currently is not quite possible. I think I set the loglevel to WARN because the RA messages I got every few minutes when I tried NOTICE.
@dedeckeh Since you made those changes to the log priorities back in 2019, maybe you can shed some light on the rationale and the best way forward?
I'll be happy to make a PR for the changes, if that's any help to you.
The text was updated successfully, but these errors were encountered: