Results of the simple experiment:
$ time makedumpfile -l -d 31 -F /proc/vmcore | cat > dump.l Copying data : [100.0 %] / eta: 0s makedumpfile Completed. real 0m0.417s #### $ time makedumpfile -c -d 31 -F /proc/vmcore | cat > dump.c Copying data : [100.0 %] | eta: 0s makedumpfile Completed. real 0m1.866s #### $ time makedumpfile -z -d 31 -F /proc/vmcore | cat > dump.z Copying data : [100.0 %] / eta: 0s makedumpfile Completed. real 0m0.802s #### $ ls -lh dump.* -rw-r--r-- 1 root root 29M Nov 1 14:04 dump.c -rw-r--r-- 1 root root 43M Nov 1 14:05 dump.l -rw-r--r-- 1 root root 26M Nov 1 14:05 dump.z -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to makedumpfile in Ubuntu. https://bugs.launchpad.net/bugs/1949395 Title: [RFE] Add zstd support for makedumpfile and use it by default Status in makedumpfile package in Ubuntu: New Bug description: Recently makedumpfile had zstd[0] support added upstream[1]. My suggestion here is to backport this work for Debian/Ubuntu and use it as default (starting for example in 22.04), since it has a better compression time and similar or even better compression rate. In [1] there are good experiments, but I did a simple experiment in a small VM (2G of RAM, idle) using Debian 11 (results in the first comment]. I'll open a Debian bug as well. Thanks in advance, Guilherme [0] https://en.wikipedia.org/wiki/Zstandard [1] http://lists.infradead.org/pipermail/kexec/2021-September/023011.html To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/makedumpfile/+bug/1949395/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
