Hi, I am starting to investigate another problem with my big installation of amanda 3.5.1 on Debian stretch, v9.
Now amanda is crashing somewhere near the end of the planning phase and the beginning of the dumping phase. In the last 4 days it happened 2 times. The information I have for now is very little. >From the report: FAILURE DUMP SUMMARY: chunker: FATAL Connection reset by peer at /usr/lib/x86_64-linux-gnu/amanda/perl/Amanda/IPC/LineProtocol.pm line 579. taper: FATAL Scribe cannot quit while a transfer is active at /usr/lib/x86_64-linux-gnu/amanda/perl/Amanda/Taper/Scribe.pm line 563. (...) >From the logs of the Linux (kern.log) I have this: Feb 22 00:47:22 XXXX kernel: warn_alloc: 222 callbacks suppressed Feb 22 00:47:22 XXXX kernel: swapper/22: page allocation failure: order:2, mode:0x2284020(GFP_ATOMIC|__GFP_COMP|__GFP_NOTRACK) Feb 22 00:47:22 XXXX kernel: CPU: 22 PID: 0 Comm: swapper/22 Not tainted 4.9.0-5-amd64 #1 Debian 4.9.65-3+deb9u2 Feb 22 00:47:22 XXXX kernel: Hardware name: Dell Inc. PowerEdge R730xd/0599V5, BIOS 2.4.3 01/17/2017 (...) Feb 22 01:11:35 XXXX kernel: traps: driver[31708] general protection ip:7fbb22b2c018 sp:7fffde984038 error:0 Feb 22 01:11:35 XXXX kernel: in libamanda-3.5.1.so[7fbb22af2000+7c000] Being that the logs of amanda have much more private information, I will send them by private email. Kind regards Jose M Calhariz -- -- "Sucesso = trabalho + estudo + boca fechada." -- Albert Einstein
