Export logs from CI in performance (preparation)

docker/test/performance-comparison/Dockerfile

@@ -56,10 +56,9 @@ COPY * /
 # node #0 should be less stable because of system interruptions. We bind
 # randomly to node 1 or 0 to gather some statistics on that. We have to bind
 # both servers and the tmpfs on which the database is stored. How to do it
-# through Yandex Sandbox API is unclear, but by default tmpfs uses
+# is unclear, but by default tmpfs uses
 # 'process allocation policy', not sure which process but hopefully the one that
-# writes to it, so just bind the downloader script as well. We could also try to
-# remount it with proper options in Sandbox task.
+# writes to it, so just bind the downloader script as well.
 # https://www.kernel.org/doc/Documentation/filesystems/tmpfs.txt
 # Double-escaped backslashes are a tribute to the engineering wonder of docker --
 # it gives '/bin/sh: 1: [bash,: not found' otherwise.

docker/test/performance-comparison/compare.sh

@@ -71,6 +71,8 @@ function configure
     while pkill -f clickhouse-serv ; do echo . ; sleep 1 ; done
     echo all killed
 
+
+
     set -m # Spawn temporary in its own process groups
 
     local setup_left_server_opts=(

docker/test/performance-comparison/entrypoint.sh

@@ -130,7 +130,7 @@ then
     git -C right/ch diff --name-only "$base" pr -- :!tests/performance :!docker/test/performance-comparison | tee other-changed-files.txt
 fi
 
-# Set python output encoding so that we can print queries with Russian letters.
+# Set python output encoding so that we can print queries with non-ASCII letters.
 export PYTHONIOENCODING=utf-8
 
 # By default, use the main comparison script from the tested package, so that we
@@ -151,11 +151,7 @@ export PATH
 export REF_PR
 export REF_SHA
 
-# Try to collect some core dumps. I've seen two patterns in Sandbox:
-# 1) |/home/zomb-sandbox/venv/bin/python /home/zomb-sandbox/client/sandbox/bin/coredumper.py %e %p %g %u %s %P %c
-#    Not sure what this script does (puts them to sandbox resources, logs some messages?),
-#    and it's not accessible from inside docker anyway.
-# 2) something like %e.%p.core.dmp. The dump should end up in the workspace directory.
+# Try to collect some core dumps.
 # At least we remove the ulimit and then try to pack some common file names into output.
 ulimit -c unlimited
 cat /proc/sys/kernel/core_pattern