Package: freedombox-setup Version: 0.8 Severity: normal -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On two separate machines, a Cubietruck and an APU 1D, I have ended up in a situation where the pam_abl database is corrupt. I assume this is after unclean shutdowns due to power outages etc. The results of this is that Apache start print the following error message in its error log: "BDB0060 PANIC: fatal region error detected; run recovery". The log keeps growing and leads to all disk space being used up. Several services then fail to start and the box is no longer accessible. The work around was to remove the pam-abl databases from /var/lib/abl/*. It appears that on unclean shutdowns, pam-able is able to corrupt its database and a manual intervention is required to clean it up. Either the upstream should use a better database than BDB (if it can't handle power cuts) or we should look for an alternative solution. - -- Sunil PS: I have no idea how Apache is related to pam-abl. Perhaps for our LDAP authentiation bits. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQIcBAEBAgAGBQJW9DjrAAoJEDbDYUQMm8lxTjcQAJX6SEv1U2qB4FC+fWvyMFHT 8UXe4N+Q+gczd/zYjaX+6Y4VCSgbBwzlh+6CCvFlXA3M56qxvS6r8OcaIHSS1q7W 36FQLt62yvLr+OuY+CjhuXQf4U57O7UKrOUAfuO1xJ0CUtdw8BaRd64OygUnOs2Q wwfue31ewcey8J1RMNgALjC7RVzU3JWx7JlHnOP9tpNAGkpnCEVtylnDyWO1yUEL S10CouAXSBQLLqP0SEUClUlZcIDsmPUr94yykuIneWAJ1jxvg7TQ5PXX4lddkt18 mV6Oyy8FVRTUbncetUe4l9CIQAJs5bLHA/YnwBz3X0gZ+3oisS+ZE/5ZmIk0+6iD nIoofjKdiNetfrF/upbGS1+hIrMDDEVHuXMrDUktblOirXSTKIRsjmCSuwjIgkVJ 6yB1JABVxkF4wH4Ban7tC3UuEKsWIjdnRLWJtxHz2/Idpfdjvl1jikG3A9m8eOKo S1sH2m1Oz6f1+qj8Fhge/lLbvwLlGsE0RcWK+xoc7HklDC0sqbB8p5zfhfUKDIJc VTJ17fVXKCaRabzcCxZTaR7Xgu3EVswt6+tJR01s5sCnHrwBFyg7EcJYK9X5bQ8F HnPiq98ao88PvsNEkHX7mKmy4aGmqb3Bfbe1sMvvKwq7rIqpChnGGIJlqobN1/no ImM8MMU7JUxp3EfecW8o =xNYB -----END PGP SIGNATURE-----
