On 2009-03-04 19:44, Dennis Peterson wrote: > Török Edwin wrote: > >> On 2009-03-04 19:28, Dennis Peterson wrote: >> > > >>> This can be tested by having some adventurous affected user run with only >>> third >>> party signatures - if it is that which is at fault then clamd will continue >>> to >>> crash. Since this happens only while loading signatures it can be tested >>> off-line. >>> >>> >> As far as I understand the crash occurs when reloading the signatures >> (actually when freeing old signatures), >> but only if clamd had some load before (i.e. it doesn't crash just by >> reloading the signatures on a fresh clamd). >> That makes me think this has to do with some memory corruption from >> earlier, or is dependent on the timing >> of events. >> >> > > Ok -- this supports the idea I have of disassociating this problem from > third-party vendors as it has bad implications for their reputations. >
The bugreport is titled "Third Party Databases: ..", because crashes have only been reported by those who use third-party databases. It is not blaming anyone, as we don't know where exactly the problem is: so far we've been unable to reproduce the crash with or without thirdparty databases. > >>> The only issue I see is that clamd CPU usage goes to 97%+ when loading a >>> new >>> signature from freshclam. >>> >>> >> That is normal. >> >> > > Is there not a "nice" way to do this? 0.95rc1 does reload "nicer", in the sense that it accepts new connections while reloading the DB. Reloading the database should take about a second. Using anything less than 100% CPU wouldn't be possible, it would just make loading the database longer. Best regards, --Edwin _______________________________________________ Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net http://www.clamav.net/support/ml
