mon/MDSMonitor.cc:refuse fs new on pools with obj #12825

stiopaa1 · 2017-01-08T20:02:47Z

Fixes: http://tracker.ceph.com/issues/11124
Signed-off-by: Michal Jarzabek stiopa@gmail.com

tchaikov · 2017-01-09T05:32:04Z

@stiopaa1 the task of test_fsnew is not referenced anywhere, you might want to add it to a new yaml file under qa/suites/fs/basic/tasks maybe? @ukernel @jcsp what do you think?

jcsp · 2017-01-09T13:13:32Z

src/mon/MDSMonitor.cc

@@ -1581,6 +1582,12 @@ int MDSMonitor::management_command(
      return -ENOENT;
    }

+    int64_t metadata_num_objects = mon->pgmon()->pg_map.pg_pool_sum[metadata].stats.sum.num_objects;


It has just occurred to me that we will sometimes want to permit this, in order to enable people to do an "fs new", "fs reset" type procedure -- that is permitted for people who make the mistake of doing an "fs rm" but still have all their objects. Could you add a --force flag here?

jcsp · 2017-01-09T13:17:10Z

qa/tasks/cephfs/test_fsnew.py

+from tasks.cephfs.cephfs_test_case import CephFSTestCase
+from teuthology.orchestra.run import CommandFailedError
+
+class TestFSNew(CephFSTestCase):


Could you move this file to test_misc.py please? That way it will get picked up and run and we avoid the need to create a separate yaml as @tchaikov mentions (it makes our tests take a bit less time if we bundle things like that, it's all a bit arbitrary though!)

jcsp · 2017-01-09T13:20:26Z

qa/tasks/cephfs/test_fsnew.py

+
+        dummyfile = '/etc/fstab'
+
+        self.fs.put_metadata_object_raw("key", dummyfile)


You could just use an inline string like "foobar" instead of dummyfile here -- put_metadata_object_raw isn't reading a file or anything like that.

I'm not sure I understand. Isn't this calling:
bin/rados -p cephfs_metadata put key "/etc/fstab"
which I thought requires a file as last argument?

jcsp · 2017-01-09T13:21:05Z

qa/tasks/cephfs/test_fsnew.py

@@ -0,0 +1,64 @@
+import errno
+import time
+from unittest import SkipTest


SkipTest import is unused

jcsp · 2017-01-09T13:21:17Z

qa/tasks/cephfs/test_fsnew.py

+
+class TestFSNew(CephFSTestCase):
+    def test_fs_new(self):
+        data_pool_name = "{0}_data".format(self.fs.name)


You can use self.fs.get_data_pool_name()

jcsp · 2017-01-09T13:22:53Z

qa/tasks/cephfs/test_fsnew.py

+        timeout = 10
+        elapsed = 0
+
+        while True:


There is a wait_until_true function for situations like this. Something like:

self.wait_until_equal(lambda: self.fs.get_pool_df(self.fs.metadata_pool_name)['objects'] > 0, timeout=timeout)

stiopaa1 · 2017-01-09T22:36:25Z

@jcsp
I've made the changes you requested (apart from 1, look at the comment above)

jcsp · 2017-01-12T00:45:39Z

Oops, you are correct about the put_metadata_object_raw usage, I had misread it

jcsp · 2017-01-13T11:52:12Z

@stiopaa1 thanks for updating, this is ready to merge, please squash

stiopaa1 · 2017-01-15T15:38:19Z

@jcsp
squashed

jcsp · 2017-01-16T11:22:44Z

Ah, now we can see that this also needs an update to the test script in qa/workunits/cephtool/test.sh:

../qa/workunits/cephtool/test.sh:911: test_mon_mds:  set +e
../qa/workunits/cephtool/test.sh:912: test_mon_mds:  ceph fs new cephfs fs_metadata mds-ec-pool
../qa/workunits/cephtool/test.sh:913: test_mon_mds:  check_response erasure-code 22 22
../qa/workunits/cephtool/test.sh:94: check_response:  expected_string=erasure-code
../qa/workunits/cephtool/test.sh:95: check_response:  retcode=22
../qa/workunits/cephtool/test.sh:96: check_response:  expected_retcode=22
../qa/workunits/cephtool/test.sh:97: check_response:  '[' 22 -a 22 '!=' 22 ']'
../qa/workunits/cephtool/test.sh:102: check_response:  grep --quiet -- erasure-code /tmp/cephtool.4bj/test_invalid.XuU
../qa/workunits/cephtool/test.sh:103: check_response:  echo 'Didn'\''t find erasure-code in output'
Didn't find erasure-code in output
../qa/workunits/cephtool/test.sh:104: check_response:  cat /tmp/cephtool.4bj/test_invalid.XuU
*** DEVELOPER MODE: setting PATH, PYTHONPATH and LD_LIBRARY_PATH ***
2017-01-16 11:20:33.164255 7fd739767700 -1 WARNING: all dangerous and experimental features are enabled.
2017-01-16 11:20:33.178010 7fd739767700 -1 WARNING: all dangerous and experimental features are enabled.
Error EINVAL: pool 'fs_metadata' already contains some objects. Use another pool which is empty.
../qa/workunits/cephtool/test.sh:105: check_response:  exit 1
../qa/workunits/cephtool/test.sh:1: check_response:  rm -fr /tmp/cephtool.4bj

You can debug this on a local vstart cluster by invoking it from a built directory something like this:
PATH=$PATH:bin/ SUDO="" ../qa/workunits/cephtool/test.sh --test-mds

stiopaa1 · 2017-01-23T00:06:20Z

@jcsp
I have updated the test script

jecluis · 2017-01-23T10:13:30Z

src/mon/MDSMonitor.cc

+    int64_t metadata_num_objects = mon->pgmon()->pg_map.pg_pool_sum[metadata].stats.sum.num_objects;
+    if (force != "--force" && metadata_num_objects > 0) {
+      ss << "pool '" << metadata_name
+	 << "' already contains some objects. Use another pool which is empty.";


a native english speaker please correct me, but this phrase feels funny. I'd personally go with "Use an empty pool instead", or something of the sorts. The which is empty portion makes me all tingly (I should probably see a doctor about that...)

It's only slightly awkward. "Use an empty pool instead" would be a bit simpler (and maybe also easier to parse for users with limited english?). Up to @stiopaa1 if you want to come back and edit this.

@jecluis @jcsp
Yeah, I agree that "Use an empty pool instead" sounds better. Will change it later today.

Fixes: http://tracker.ceph.com/issues/11124 Signed-off-by: Michal Jarzabek <stiopa@gmail.com>

stiopaa1 · 2017-01-23T19:54:41Z

updated to "Use an empty pool instead"

tchaikov added the cephfs Ceph File System label Jan 9, 2017

jcsp reviewed Jan 9, 2017

View reviewed changes

qa/tasks/cephfs/test_fsnew.py

@@ -0,0 +1,64 @@

import errno

import time

from unittest import SkipTest

Copy link

Contributor

jcsp Jan 9, 2017

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

SkipTest import is unused

jcsp reviewed Jan 9, 2017

View reviewed changes

jcsp approved these changes Jan 13, 2017

View reviewed changes

stiopaa1 force-pushed the r11124_refuseToDoFsNew branch from 59cc219 to 7942dc7 Compare January 15, 2017 09:43

stiopaa1 force-pushed the r11124_refuseToDoFsNew branch from 7942dc7 to 21030ff Compare January 22, 2017 22:54

jecluis reviewed Jan 23, 2017

View reviewed changes

mon/MDSMonitor.cc:refuse fs new on pools with obj

052c3d3

Fixes: http://tracker.ceph.com/issues/11124 Signed-off-by: Michal Jarzabek <stiopa@gmail.com>

stiopaa1 force-pushed the r11124_refuseToDoFsNew branch from 21030ff to 052c3d3 Compare January 23, 2017 19:52

jcsp merged commit 48adc77 into ceph:master Jan 27, 2017

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

mon/MDSMonitor.cc:refuse fs new on pools with obj #12825

mon/MDSMonitor.cc:refuse fs new on pools with obj #12825

stiopaa1 commented Jan 8, 2017

tchaikov commented Jan 9, 2017 •

edited

jcsp Jan 9, 2017

jcsp Jan 9, 2017

jcsp Jan 9, 2017

stiopaa1 Jan 9, 2017

jcsp Jan 9, 2017

jcsp Jan 9, 2017

jcsp Jan 9, 2017

stiopaa1 commented Jan 9, 2017

jcsp commented Jan 12, 2017

jcsp commented Jan 13, 2017

stiopaa1 commented Jan 15, 2017

jcsp commented Jan 16, 2017

stiopaa1 commented Jan 23, 2017

jecluis Jan 23, 2017

jcsp Jan 23, 2017

stiopaa1 Jan 23, 2017

stiopaa1 commented Jan 23, 2017 •

edited


		dummyfile = '/etc/fstab'

		self.fs.put_metadata_object_raw("key", dummyfile)

mon/MDSMonitor.cc:refuse fs new on pools with obj #12825

mon/MDSMonitor.cc:refuse fs new on pools with obj #12825

Conversation

stiopaa1 commented Jan 8, 2017

tchaikov commented Jan 9, 2017 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stiopaa1 commented Jan 9, 2017

jcsp commented Jan 12, 2017

jcsp commented Jan 13, 2017

stiopaa1 commented Jan 15, 2017

jcsp commented Jan 16, 2017

stiopaa1 commented Jan 23, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stiopaa1 commented Jan 23, 2017 • edited

tchaikov commented Jan 9, 2017 •

edited

stiopaa1 commented Jan 23, 2017 •

edited