]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Sage Weil [Tue, 10 Nov 2009 05:31:51 +0000 (21:31 -0800)]
mon: ignore MAuth message when connection has no session
This happens if hte message is delayed, and the session meanwhile
is closed. We need similar checks any time we take a Session
from the message->get_connection()....
Sage Weil [Tue, 10 Nov 2009 05:30:11 +0000 (21:30 -0800)]
ceph: default to admin user
Sage Weil [Tue, 10 Nov 2009 05:30:03 +0000 (21:30 -0800)]
auth: make ~ in keyring search path subst in $HOME
Sage Weil [Tue, 10 Nov 2009 05:12:16 +0000 (21:12 -0800)]
Merge branch 'unstable' into auth
Conflicts:
src/Makefile.am
src/mkmonfs.cc
src/msg/SimpleMessenger.cc
Yehuda Sadeh [Tue, 10 Nov 2009 00:54:08 +0000 (16:54 -0800)]
auth: auth none encodes global_id at build_authorizer
Yehuda Sadeh [Tue, 10 Nov 2009 00:31:00 +0000 (16:31 -0800)]
auth: pass global_id to the osd, mds verify-authorizers
Sage Weil [Tue, 10 Nov 2009 00:08:50 +0000 (16:08 -0800)]
mds, osd: call monc->authenticate(), not wait_authenticate()
Sage Weil [Tue, 10 Nov 2009 00:07:59 +0000 (16:07 -0800)]
uclient: get keys, build authorizer
Yehuda Sadeh [Mon, 9 Nov 2009 23:43:53 +0000 (15:43 -0800)]
auth: don't display auth incrementals in the ceph tool
Sage Weil [Mon, 9 Nov 2009 23:52:11 +0000 (15:52 -0800)]
auth: fix global id alloc on peon monitors
Sage Weil [Mon, 9 Nov 2009 23:38:56 +0000 (15:38 -0800)]
auth: kill ClientMOnitor, MCLientMOunt*, etc.
Sage Weil [Mon, 9 Nov 2009 22:31:55 +0000 (14:31 -0800)]
mon: use g_conf.mon_globalid_prealloc
Yehuda Sadeh [Mon, 9 Nov 2009 23:00:49 +0000 (15:00 -0800)]
auth: added should_propose callback to auth mon
Yehuda Sadeh [Mon, 9 Nov 2009 22:02:11 +0000 (14:02 -0800)]
auth: verify global_id with the one on the ticket
Yehuda Sadeh [Mon, 9 Nov 2009 22:01:43 +0000 (14:01 -0800)]
auth: increase max global id only when paxos is writable
Sage Weil [Mon, 9 Nov 2009 21:17:29 +0000 (13:17 -0800)]
osd: log misdirected ops; reply with -ENXIO
This is more helpful than assert(0). It's still bad (it means the client
and osd calculated different pg mappings) though, but this makes it easier
to identify and fix.
Sage Weil [Mon, 9 Nov 2009 21:01:08 +0000 (13:01 -0800)]
sample.ceph.conf: fix btrfs mountoptions
Yehuda Sadeh [Mon, 9 Nov 2009 18:16:04 +0000 (10:16 -0800)]
auth: distribute global instance id to client
Sage Weil [Sun, 8 Nov 2009 17:21:16 +0000 (09:21 -0800)]
osdmap: clear out old hash distribution code
This screws up linkage because not everything that #Includes osdmap.h
links crush.
Sage Weil [Sun, 8 Nov 2009 05:06:27 +0000 (21:06 -0800)]
osd: make pgids sort on (pool, preferred, ps)
This makes pg dump output easier to read, mainly.
Sage Weil [Sun, 8 Nov 2009 05:05:32 +0000 (21:05 -0800)]
crushtool: small fix
Sage Weil [Sun, 8 Nov 2009 04:49:37 +0000 (20:49 -0800)]
Revert "crush: use spirit classic headers"
This reverts commit
28ac4441b87907713ddaf1fe1dee62350f947cf3 .
Sage Weil [Sat, 7 Nov 2009 23:27:08 +0000 (15:27 -0800)]
hash: small cleanup
Sage Weil [Sun, 8 Nov 2009 04:12:21 +0000 (20:12 -0800)]
crush: hrm fix up builder too
fix
Sage Weil [Sun, 8 Nov 2009 03:59:42 +0000 (19:59 -0800)]
crush: use spirit classic headers
This makes the 'deprecated' warnings go away.
Sage Weil [Sun, 8 Nov 2009 03:59:05 +0000 (19:59 -0800)]
crush: make hash function selectable
Sage Weil [Sat, 7 Nov 2009 05:36:43 +0000 (21:36 -0800)]
osd: make object hash a pg_pool parameter
Sage Weil [Sat, 7 Nov 2009 00:48:38 +0000 (16:48 -0800)]
osdmaptool: test-map-object
Sage Weil [Sat, 7 Nov 2009 00:43:47 +0000 (16:43 -0800)]
osd: use stronger hash function for mapping objects -> pgs
The old hash (from linux dcache) was very weak, such that
least sig bits may not change and you could get lots of
consecutive objects on the same osds (because lsbits of the
pg weren't changing).
This is Robert Jenkin's hash and is quite strong. Public
domain.
Rev osd disk format, protocol, since we're totally changing
object placement here.
Sage Weil [Fri, 6 Nov 2009 23:37:01 +0000 (15:37 -0800)]
crush: no more static inline
Yehuda Sadeh [Sat, 7 Nov 2009 00:43:21 +0000 (16:43 -0800)]
auth: some cleanup, compilation fixes
Yehuda Sadeh [Fri, 6 Nov 2009 22:30:38 +0000 (14:30 -0800)]
auth: manage global id in the auth mon
Greg Farnum [Fri, 6 Nov 2009 23:27:23 +0000 (15:27 -0800)]
osd: This logic is slightly less confusing without the always-true 'full' param
Sage Weil [Fri, 6 Nov 2009 21:54:53 +0000 (13:54 -0800)]
mon: don't delete stats when sending osdmap incremental
send_latest will delete m, and/or wait. Instead call send_incremental
directly only when we know paxos is_readable.
Sage Weil [Fri, 6 Nov 2009 20:12:24 +0000 (12:12 -0800)]
qa: test subdir mounts
Sage Weil [Fri, 6 Nov 2009 06:22:27 +0000 (22:22 -0800)]
init-ceph: tell user we're mounting btrfs
Sage Weil [Fri, 6 Nov 2009 21:36:32 +0000 (13:36 -0800)]
osd: send single osd_boot on startup
Sage Weil [Fri, 6 Nov 2009 21:36:25 +0000 (13:36 -0800)]
mon: don't log dup osd boot msgs
Sage Weil [Fri, 6 Nov 2009 21:35:08 +0000 (13:35 -0800)]
msgr: leave off ss_family when printing
Sage Weil [Fri, 6 Nov 2009 06:21:22 +0000 (22:21 -0800)]
todo
Sage Weil [Fri, 6 Nov 2009 21:09:58 +0000 (13:09 -0800)]
ppc: do not copy_in unencoded __u32
Fixes big endian bugs.
Sage Weil [Fri, 6 Nov 2009 21:09:26 +0000 (13:09 -0800)]
buffer: only define _XOPEN_SOURCE ifndef
Sage Weil [Fri, 6 Nov 2009 07:03:59 +0000 (23:03 -0800)]
todo
Sage Weil [Fri, 6 Nov 2009 06:25:30 +0000 (22:25 -0800)]
mon: make initial monmap epoch match paxos version
Sage Weil [Fri, 6 Nov 2009 06:24:19 +0000 (22:24 -0800)]
mkmonfs: use common_init and parse regular args
Sage Weil [Fri, 6 Nov 2009 05:53:15 +0000 (21:53 -0800)]
don't forget -standalone.git in release checklist
Greg Farnum [Thu, 5 Nov 2009 23:58:27 +0000 (15:58 -0800)]
OSD: Don't repeatedly send the same incremental map update to clients.
Greg Farnum [Wed, 4 Nov 2009 00:43:53 +0000 (16:43 -0800)]
uclient: set NAME_MAX = PAGE_SIZE
Greg Farnum [Thu, 5 Nov 2009 20:58:29 +0000 (12:58 -0800)]
Hadoop: Update JavaDoc and put in new patch file
Greg Farnum [Wed, 4 Nov 2009 22:14:41 +0000 (14:14 -0800)]
Hadoop: Numerous fixes.
Set bufUsed=0 on a flush to avoid bad rewrites of data
Downgrade a warning in IOStream since Hadoop apparently checks for EOF by reading until a read returns -1.
Remove some leftover if checks that don't do anything.
TODO: Remove something Sage did already.
Yehuda Sadeh [Thu, 5 Nov 2009 18:00:19 +0000 (10:00 -0800)]
mount: fix hint initialization for getaddrinfo
Yehuda Sadeh [Thu, 5 Nov 2009 18:00:19 +0000 (10:00 -0800)]
mount: fix hint initialization for getaddrinfo
Sage Weil [Thu, 5 Nov 2009 05:43:47 +0000 (21:43 -0800)]
filestore: oops less noisy
Sage Weil [Thu, 5 Nov 2009 00:21:20 +0000 (16:21 -0800)]
osdmap: fix pps calc for preferred pgs
Sage Weil [Thu, 5 Nov 2009 00:09:32 +0000 (16:09 -0800)]
osd: nicer debug output on misdirected requests
Sage Weil [Wed, 4 Nov 2009 23:22:54 +0000 (15:22 -0800)]
mds: mark scatterlocks with dirty rstat/fragstat during replay
This ensures we propagate this info back toward the root
after we've replayed and gone active.
Sage Weil [Wed, 4 Nov 2009 23:22:11 +0000 (15:22 -0800)]
mds: use fnode when replaying journal
We weren't actually pulling the fnode from the journal. I'm
surprised anything worked.
This fixes a crash when untaring, restarting mds, and then
rm -r'ing the dir. Previously the rstat would roll
negative and assert.
Sage Weil [Wed, 4 Nov 2009 21:27:06 +0000 (13:27 -0800)]
debian: gracefully replace lib packages prior to '1' suffix
Sage Weil [Wed, 4 Nov 2009 20:24:39 +0000 (12:24 -0800)]
osd: kill int <-> pg_t conversions
These are messy and asking for trouble. And a cleaner
coll_t paves the way for real named coll_t's down the
line.
Sage Weil [Wed, 4 Nov 2009 19:37:41 +0000 (11:37 -0800)]
osd: convert ceph_pg union to struct
This simplifies/fixes endian conversions.
Sage Weil [Tue, 3 Nov 2009 23:15:22 +0000 (15:15 -0800)]
msgr: encode sockaddr.ss_family big endian in ceph_entity_addr
The ss_family field is normally host endianness, but we
want to exchange ceph_entity_addr across the wire and store
it on disk. So, encode ss_family in big endian (to match
the other sockaddr field endianness).
Rev disk and wire protocols to match.
Sage Weil [Tue, 3 Nov 2009 20:31:40 +0000 (12:31 -0800)]
filestore: close btrfs transactions cleanly
Don't just close the fd. We may add a mount option that makes a messy
transaction close wedge the fs instead of committing partial results.
Sage Weil [Tue, 27 Oct 2009 15:39:02 +0000 (08:39 -0700)]
default pid file /var/run/ceph/$type.$id.pid
Yehuda Sadeh [Tue, 3 Nov 2009 20:23:49 +0000 (12:23 -0800)]
auth: fix authorizer in the 'none' protocol case
Yehuda Sadeh [Tue, 3 Nov 2009 20:23:34 +0000 (12:23 -0800)]
vstart.sh: add option to avoid conf file overwrite
Yehuda Sadeh [Tue, 3 Nov 2009 18:38:50 +0000 (10:38 -0800)]
auth: set caps to allow-all with auth-none
Greg Farnum [Tue, 3 Nov 2009 00:37:38 +0000 (16:37 -0800)]
Hadoop: Don't use the local pg by default.
Greg Farnum [Mon, 2 Nov 2009 04:18:29 +0000 (20:18 -0800)]
Now passes all unit tests!
Changes made to rename and between error codes/exception throwing.
Greg Farnum [Sat, 31 Oct 2009 01:39:48 +0000 (18:39 -0700)]
Hadoop: Don't throw IOExceptions on extra calls to close
Greg Farnum [Fri, 30 Oct 2009 22:37:06 +0000 (15:37 -0700)]
Hadoop: f5 e3
Greg Farnum [Fri, 30 Oct 2009 20:53:54 +0000 (13:53 -0700)]
Hadoop: Behavioral fixes to CephFileSystem
Greg Farnum [Fri, 30 Oct 2009 20:50:30 +0000 (13:50 -0700)]
Hadoop: Various changes to CephFaker; not completed.
Greg Farnum [Thu, 29 Oct 2009 22:06:12 +0000 (15:06 -0700)]
HadooP: CephFaker now properly sanitizes input filenames.
Greg Farnum [Thu, 29 Oct 2009 01:36:22 +0000 (18:36 -0700)]
Hadoop: Update constructor/initialize usage a bit
Greg Farnum [Wed, 28 Oct 2009 21:04:58 +0000 (14:04 -0700)]
Simplify CephInputStream so that it actually works
Greg Farnum [Tue, 27 Oct 2009 20:57:29 +0000 (13:57 -0700)]
Hadoop: Simplify CephOutputStream so that it actually works.
Greg Farnum [Tue, 27 Oct 2009 20:02:30 +0000 (13:02 -0700)]
Hadoop: Reorder a few things for better safety and fix compile bugs
Yehuda Sadeh [Mon, 2 Nov 2009 22:52:34 +0000 (14:52 -0800)]
auth: don't wait for rotating secret with auth-none
Sage Weil [Mon, 2 Nov 2009 01:51:51 +0000 (17:51 -0800)]
crush: always return a value from crush_bucket_choose
Yehuda Sadeh [Fri, 30 Oct 2009 23:17:57 +0000 (16:17 -0700)]
auth: fix 'auth none' authentication
Sage Weil [Fri, 30 Oct 2009 22:13:44 +0000 (15:13 -0700)]
mds: set isvalid in verify_authroizer
Yehuda Sadeh [Fri, 30 Oct 2009 21:48:17 +0000 (14:48 -0700)]
auth: fix osd authorizer
Yehuda Sadeh [Fri, 30 Oct 2009 21:18:54 +0000 (14:18 -0700)]
auth: add mds authorizer
Sage Weil [Fri, 30 Oct 2009 20:49:04 +0000 (13:49 -0700)]
auth: search list of files when loading keyring
Yehuda Sadeh [Fri, 30 Oct 2009 21:02:56 +0000 (14:02 -0700)]
auth: allocate authorizers once
Yehuda Sadeh [Fri, 30 Oct 2009 20:20:53 +0000 (13:20 -0700)]
auth: define authorize callbacks for different auth protocols
Sage Weil [Fri, 30 Oct 2009 19:38:17 +0000 (12:38 -0700)]
auth: always verify auth reply, even if it's length 0
A length 0 reply is probably invalid, meaning we didn't
successfully authenticate our peer!
Yehuda Sadeh [Fri, 30 Oct 2009 17:09:35 +0000 (10:09 -0700)]
include guards cleanup
Yehuda Sadeh [Thu, 29 Oct 2009 23:40:41 +0000 (16:40 -0700)]
auth: less chatty
Yehuda Sadeh [Thu, 29 Oct 2009 23:35:13 +0000 (16:35 -0700)]
auth: configuratble supported auth type
Sage Weil [Thu, 29 Oct 2009 23:24:40 +0000 (16:24 -0700)]
todo
Sage Weil [Thu, 29 Oct 2009 22:49:09 +0000 (15:49 -0700)]
auth: fix verify_authorize_reply stupid
Sage Weil [Thu, 29 Oct 2009 22:33:22 +0000 (15:33 -0700)]
auth: use u64 nonce, not timestamp, in authorizer[_reply]
Sage Weil [Thu, 29 Oct 2009 22:33:08 +0000 (15:33 -0700)]
auth: fix service side bug on bad authorizer
Sage Weil [Thu, 29 Oct 2009 22:32:52 +0000 (15:32 -0700)]
auth: AuthBlob -> CephXTicketBlob
Sage Weil [Thu, 29 Oct 2009 20:28:09 +0000 (13:28 -0700)]
auth: simplify challenge key encoding
Sage Weil [Thu, 29 Oct 2009 20:11:43 +0000 (13:11 -0700)]
buffer: make [], hexdump const
Sage Weil [Thu, 29 Oct 2009 20:11:27 +0000 (13:11 -0700)]
auth: break client/server challenge calc into helper
Yehuda Sadeh [Thu, 29 Oct 2009 22:03:51 +0000 (15:03 -0700)]
auth: CEPH_AUTH_NONE auth and caps working
Yehuda Sadeh [Thu, 29 Oct 2009 19:34:36 +0000 (12:34 -0700)]
auth: return correct auth type