In "direct zope access" ports, the shared frontend is not used, so
the argument that long timeout consume resources on shared server does
not apply here.

A timeout of one hour was choosen arbitrarily, a value that should be
large enough for normal requests and more than the default 60s timeout
that we hit in the "wait for activities" step when running zelenium
tests.

quoting https://en.wikipedia.org/wiki/Slowloris_(computer_security) :

> Since Apache 2.2.15, Apache ships the module mod_reqtimeout as the
> official solution supported by the developers

From https://httpd.apache.org/docs/2.4/mod/mod_reqtimeout.html
default is header=20-40,MinRate=500 body=20,MinRate=500
which seem to be good default.

https://bugs.launchpad.net/zope2/+bug/699865

usually, the software release path is :
in a web runner : /srv/slapgrid/slappartXX/srv/runner/project/buildout.cfg
in a test node : /srv/slapgrid/slappartX/srv/testnode/xx/inst/test0-XX/srv/runner/project/buldout.cfg

the previous regex was replacing in all slappartX and test0-X by the new test0-Y
which of course would make an existing path : /srv/slapgrid/test0-Y/.../test0-Y/...

This commit only replace the latest occurence in order to have
compatible code in the "normal" case, and in the test nodes

/reviewed-on nexedi/slapos!418

/reviewed-on nexedi/slapos!414

These are totally unrelated changes which were not meant to be pushed,
sorry.

Runner exporter failed with this message while checking if file changed during backup of instances files:

    < 3226bab4cd318d30cfbde18e6530435dcae3205f94103384ce266c3c73d1afeb  ./runner/instance/slappart2/srv/backup/pbs/slaprunner-slappart1-runner-1/backup.signature
    < 3226bab4cd318d30cfbde18e6530435dcae3205f94103384ce266c3c73d1afeb  ./runner/instance/slappart3/srv/backup/runner/backup.signature
    < 3226bab4cd318d30cfbde18e6530435dcae3205f94103384ce266c3c73d1afeb  ./runner/instance/slappart5/srv/backup/runner/backup.signature
    5248d5244
    < bb3d96876d1d6cecee9878d835e7515215fc5d2a6ae6b647d0e7915c36700b54  ./runner/instance/slappart4/srv/backup/runner/backup.signature
    ERROR: Some backups are not consistent, exporter should be re-run.

This is because instance in webrunner has also backup.signature the file is excluded while doing sha256sum on backup file:

    find -type f ! -name backup.signature -print0 | xargs -0 sha256sum | sort -k 66 > backup.signature

A fix is to exclude backup.signature in `find`.

If instance/slappartXX is not a folder but file, skip chdir else import/export fail with a message like:

    /srv/slapgrid/slappart18/bin/runner-exporter run at : Thu Oct  4 04:30:03 CEST 2018
    Traceback (most recent call last):
      File "<stdin>", line 17, in <module>
    OSError: [Errno 20] Not a directory: '/srv/slapgrid/slappart18/srv//runner/instance/slappart5-mariadb.dump.sql.gz'

This reverts commit 80673c8c.

All testnodes should have been updated by now, so this commit is no
longer needed.

/reviewed-on nexedi/slapos!419

/reviewed-on nexedi/slapos!416

/reviewed-on nexedi/slapos!415

/reviewed-on nexedi/slapos!408

isp data is from https://www.whoismyisp.org/ip
cn-t for china telecom, default one if isp is other value
cn-m for china mobile
cn-u for china unicom

fix rst syntax - second try

fix markdown syntax

from nexedi/slapos!408 (comment 66703)

as always, using wrapper is easier and more flexible

also use a real bash and not busybox

   bzip.org don't exist anymore, download from fosseis is ok as the md5sum guarantee we still get the same file.

/reviewed-on nexedi/slapos!406

LIBPCRE1_JIT is required to compile git-2.17.1, so we enabled jit in pcre component. Another solution will be to compile git with NO_LIBPCRE1_JIT=1 and keep jit disabled in pcre.