Log display / backup fails

Accounts

2.0-ALPHA-ALPHA built on Sun Mar 22 00:52:17 EDT 2009 Current snapshot

logs fail to display via www interface might have to do with this error.
"Warning: date() expects parameter 2 to be long, string given in /usr/local/www/guiconfig.inc on line 767"

and I can not backup my config file, Endless waiting for download under IE and also firefox.

Also NAT redirect seems to be failing as I can not access some sites on my other public ip's via 1:1 nat.

tohil

at my installation, log works not correctly

2.0-ALPHA-ALPHA
built on Sun Mar 22 06:10:09 EDT 2009
FreeBSD 7.1-RELEASE-p3

the firewall log is not updating clean, and sometimes not the last 50 entries are shown if i refresh. there are shown just ~10 entries or something like that.

i've now made a ssh, telnet, 3389, icmp ping to my firewall from an external system. but the log is not updating. ive log everything.

the last log entry is from Mar 24 11:09:20…. but now its 11:26. still no entry after 2minutes after the blocked access

nocer

Hi,

fifolog screwed up entire logging unfortunately. can't be obtained any logs from the gui, gives core while trying to read logs by fifolog_reader command at the console. It's really hard to tell what is happening to my box without having event logging usable.

Would you developers really mind to fix it please.

cheers,

2.0-ALPHA-ALPHA
built on Wed Mar 25 21:57:22 EDT 2009
FreeBSD 7.1-RELEASE-p4

databeestje

Well, we used clog previously which started core dumping and no easy solution was found. The other problem was that it was unmaintained for the last few years.

Since fifolog is part of FreeBSD 7.1 we wanted to use something which is actively maintained.

We are kind of caught off guard by the fact that it is so wildly unstable. We are currently evaluating our options.

jimp

Some changes were made last night by sullrich/GeekGod to bring clog back into the mix, but also to make logging a little more generalized, so there may be options for clog and fifolog both in the long run.

It should be back to normal with a snapshot from today, but I haven't tested it yet.