I built and installed a new Squid (2.5-Stable1) to replace the existing 2.4-Stable7 I had been running. The build options were the same, but when I ran 2.5 it wouldn't stay up. The following appeared in my logs:
Mar 19 13:59:44 squid[53149]: Store rebuilding is 1.3% complete Mar 19 14:00:27 squid[53149]: xcalloc: Unable to allocate 1 blocks of 36 bytes! Mar 19 14:00:27 squid[53149]: Not currently OK to rewrite swap log. Mar 19 14:00:27 squid[53149]: storeDirWriteCleanLogs: Operation aborted. Mar 19 14:00:28 /kernel: pid 53149 (squid), uid 65534: exited on signal 6 ..and then RunCache restarts it, and the process continues in this cycle. I've been running 2.4-Stable7 (and other previous versions) on the same box (FreeBSD 4.x, currently 4.8-RC) for a couple of years without any issues, including building new squids. Is there something about 2.5 that makes it not like a 2.4-compatible cache? Is my cache corrupt? Anything that can be done other than purging the cache_dir?
