sapling/tests/test-infinitepushbackup-sql1.t
Mark Thomas 5180dd5cf2 infinitepush: split backups into infinitepushbackup extension
Summary:
Split the infinitepush backups code into a separate module.

The infinitepush extension is responsible for providing scratch pushes, as well
as the server-side implementation for storing scratch bundles.

The infinitepushbackup extension uses this feature to provide automatic backup
support.  It is an alternative to the commitcloud extension, and the two will
be mutually exclusive.

Reviewed By: DurhamG

Differential Revision: D8204418

fbshipit-source-id: 9279c426da35d451a5e7b39a72f2c91d364e3fd4
2018-06-05 04:36:06 -07:00

214 lines
7.5 KiB
Perl

#if no-windows no-osx
$ mkcommit() {
> echo "$1" > "$1"
> hg add "$1"
> hg ci -d "0 0" -m "$1"
> }
$ . "$TESTDIR/infinitepush/library.sh"
$ setupcommon
With no configuration it should abort
$ hg init server
$ cd server
$ setupsqlserverhgrc babar
$ hg st
abort: please set infinitepush.sqlhost
[255]
$ setupdb
$ cd ..
$ hg clone -q ssh://user@dummy/server client1
$ hg clone -q ssh://user@dummy/server client2
$ cd client1
$ setupsqlclienthgrc
$ cd ../client2
$ setupsqlclienthgrc
$ cd ../client1
$ mkcommit scratchcommit
$ hg push -r . --to scratch/book --create
pushing to ssh://user@dummy/server
searching for changes
remote: pushing 1 commit:
remote: 2d9cfa751213 scratchcommit
Make pull and check that scratch commit is not pulled
$ cd ../client2
$ hg pull
pulling from ssh://user@dummy/server
no changes found
$ hg log -r scratch/book
abort: unknown revision 'scratch/book'!
[255]
Pull scratch commit from the second client
$ hg pull -B scratch/book
pulling from ssh://user@dummy/server
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files
new changesets 2d9cfa751213
(run 'hg update' to get a working copy)
$ hg up scratch/book
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
(activating bookmark scratch/book)
$ hg log -G
@ changeset: 0:2d9cfa751213
bookmark: scratch/book
tag: tip
user: test
date: Thu Jan 01 00:00:00 1970 +0000
summary: scratchcommit
$ cd ../server
$ hg book scratch/%erversidebook
$ hg book serversidebook
$ cd ../client1
$ hg book --list-remote 'scratch/*'
scratch/%erversidebook 0000000000000000000000000000000000000000
scratch/book 2d9cfa7512136a84a6edb6a7c288145229c2ef7f
$ hg book --list-remote 'scratch/%*'
scratch/%erversidebook 0000000000000000000000000000000000000000
$ mysql -h $DBHOST -P $DBPORT -D $DBNAME -u $DBUSER $DBPASSOPT -e 'select * from nodesmetadata'
node message p1 p2 author committer author_date committer_date reponame optional_json_metadata
2d9cfa7512136a84a6edb6a7c288145229c2ef7f scratchcommit 0000000000000000000000000000000000000000 0000000000000000000000000000000000000000 test test 0 0 babar NULL
$ cd ../server
$ hg debugfillinfinitepushmetadata --node 2d9cfa7512136a84a6edb6a7c288145229c2ef7f
$ mysql -h $DBHOST -P $DBPORT -D $DBNAME -u $DBUSER $DBPASSOPT -e 'select * from nodesmetadata'
node message p1 p2 author committer author_date committer_date reponame optional_json_metadata
2d9cfa7512136a84a6edb6a7c288145229c2ef7f scratchcommit 0000000000000000000000000000000000000000 0000000000000000000000000000000000000000 test test 0 0 babar {"changed_files": {"scratchcommit": {"adds": 1, "isbinary": false, "removes": 0, "status": "added"}}}
$ cd ../
$ rm -rf client1 && rm -rf client2
Test ordering of hosts and reporoot in pullbackup
Expected result:
repo roots must be ordered in MRU order (the way the the most recent used come first)
Since bookmarkstonode mysql table uses '`time` datetime DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP'
with second resolution instead of an AUTO_INCREMENT column,
sleeps should be used to get predictable order
$ hg clone -q ssh://user@dummy/server client1
$ cd client1 && setupsqlclienthgrc
$ mkcommit 'Commit in repo client1'
$ sleep 1
$ hg pushb -q --config infinitepushbackup.hostname=mydevhost
$ cd ..
$ hg clone -q ssh://user@dummy/server 'client2\dev'
$ cd 'client2\dev' && setupsqlclienthgrc
$ mkcommit 'Commit in repo client2\dev'
$ sleep 1
$ hg pushb -q --config infinitepushbackup.hostname=devhost
$ cd ..
$ hg clone -q ssh://user@dummy/server client3
$ cd client3 && setupsqlclienthgrc
$ mkcommit 'Commit in repo client3'
$ sleep 1
$ hg pushb -q --config infinitepushbackup.hostname=devhost
$ cd ..
$ hg clone -q ssh://user@dummy/server client4
$ cd client4 && setupsqlclienthgrc
$ mkcommit 'Commit in repo client4'
$ sleep 1
$ hg pushb -q --config infinitepushbackup.hostname=ourdevhost
$ cd ..
$ hg clone -q ssh://user@dummy/server client5
$ cd client5 && setupsqlclienthgrc
$ mkcommit 'Commit in repo client5'
$ sleep 1
$ hg pushb -q --config infinitepushbackup.hostname=mydevhost
$ hg pullbackup --config infinitepushbackup.backuplistlimit=3
user test has 5 available backups:
(backups are ordered with the most recent at the top of the list)
$TESTTMP/client5 on mydevhost
$TESTTMP/client4 on ourdevhost
$TESTTMP/client3 on devhost
(older backups have been hidden, run 'hg getavailablebackups --all' to see them all)
abort: multiple backups found
(set --hostname and --reporoot to pick a backup)
[255]
$ hg pullbackup --hostname mydevhost
user test has 2 available backups:
(backups are ordered with the most recent at the top of the list)
$TESTTMP/client5 on mydevhost
$TESTTMP/client1 on mydevhost
abort: multiple backups found
(set --hostname and --reporoot to pick a backup)
[255]
$ hg pullbackup --reporoot $TESTTMP/client1 --hostname mydevhost -q # pullbackup should work with reporoot/hostname pair
$ hg pullbackup --reporoot $TESTTMP/'client2\dev' --hostname devhost -q # pullbackup should work with reporoot/hostname pair
$ hg pullbackup --reporoot $TESTTMP/client3 -q # pullbackup should work with reporoot only if it is unambiguous
$ hg pullbackup --hostname ourdevhost -q # pullbackup should work with hostname only if it is unambiguous
$ hg log -G --template "{node|short} '{desc}'\n"
o 3309f3c00117 'Commit in repo client4'
o 8b99f4b01a41 'Commit in repo client3'
o f9fdbb60b59e 'Commit in repo client2\dev'
o 250e5455acab 'Commit in repo client1'
@ ec33790d6769 'Commit in repo client5'
Getavailablebackups should also go in MRU order
$ hg getavailablebackups --json
{
"mydevhost": [
"$TESTTMP/client5",
"$TESTTMP/client1"
],
"ourdevhost": [
"$TESTTMP/client4"
],
"devhost": [
"$TESTTMP/client3",
"$TESTTMP/client2\\dev"
]
}
$ hg getavailablebackups
user test has 5 available backups:
(backups are ordered with the most recent at the top of the list)
$TESTTMP/client5 on mydevhost
$TESTTMP/client4 on ourdevhost
$TESTTMP/client3 on devhost
$TESTTMP/client2\dev on devhost
$TESTTMP/client1 on mydevhost
$ cd ../
$ rm -rf client*
Test of creation and later removal a bookmark that ends with '*'
Extected that the bookmark 'feature1' will mot be removed and
safely will be restored on the client2
$ hg clone -q ssh://user@dummy/server client1
$ hg clone -q ssh://user@dummy/server client2
$ cd client1 && setupsqlclienthgrc && cd ..
$ cd client2 && setupsqlclienthgrc && cd ..
$ cd client1
$ mkcommit 'Test commit1'
$ hg bookmark 'feature1'
$ sleep 1
$ mkcommit 'Test commit2'
$ hg up '.^'
0 files updated, 0 files merged, 1 files removed, 0 files unresolved
(leaving bookmark feature1)
$ hg bookmark '*'
$ sleep 1
$ hg log -G --template "{node|short} {bookmarks} '{desc}'\n"
o 679ff862f673 feature1 'Test commit2'
|
@ bf45ce9ee037 * 'Test commit1'
$ hg pushb -q --config infinitepushbackup.hostname=testhost
$ hg bookmark '*' --delete
$ hg pushb -q --config infinitepushbackup.hostname=testhost
$ cd ../client2
$ hg pullbackup --hostname testhost -q
$ hg log -G --template "{node|short} {bookmarks} '{desc}'\n"
o 679ff862f673 feature1 'Test commit2'
|
o bf45ce9ee037 'Test commit1'
#endif