Usually it can get it from the OSD, but it assumes the mon will also share
and doesn't behave when it does not.
Fixes: #5767
Signed-off-by: Sage Weil <sage@inktank.com>
It's a wad of special cases, but it implements "tell <pgid>" such that
it has the same effect as "pg <pgid>".
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Contact an OSD that's up to get a list of the commands, and use
them to add to the URL map.
Special treatment throughout for these commands:
* hack the help signature dump
* keep a 'flavor' per command to allow special handler() processing
* strip off 'tell/<target>' when constructing command
* allow multiple dicts with the same url
(the parameters and get/put methods can change)
* because of above, method must be validated in handler()
* validate the given OSD
* calculate target for command (mon, osd, pg)
Unrelated: make method_dict into global METHOD_DICT
Signed-off-by: Dan Mick <dan.mick@inktank.com>
We might have a sequence like:
- start mon, load pgmap 100
- sync
- including a format upgrade at say v 150
- refresh
- see format_version==1, and try read pgmap:101 as new format
This simply clears our in-memory state if we see that the format has
changed. That will make update_from_paxos reload the latest and prevent
it from walking through the old and useless inc updates.
Note: this does not affect the auth monitor because we unconditionally
load the latest map in update_from_paxos on upgrade. Also, the upgrade
there wasn't a format change--just a translation of cap strings from the
old to new style.
Fixes: #5764
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com>
Call delete on md_log in the destructor.
CID 1054826 (#1 of 1): Resource leak in object (CTOR_DTOR_LEAK)
1. alloc_new: Allocating memory by calling "new RGWMetadataLog(_cct, _store)".
2. var_assign: Assigning: "this->md_log" = "new RGWMetadataLog(_cct, _store)".
3. ctor_dtor_leak: The constructor allocates field "md_log" of
"RGWMetadataManager" but the destructor and whatever functions it calls
do not free it.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Call complete_list_entries() at end of RGWOp_MDLog_List::execute()
to free the 'handle'.
CID 1054834 (#1-2 of 2): Resource leak (RESOURCE_LEAK)
leaked_storage: Variable "handle" going out of scope leaks the
storage it points to.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Fixes for:
CID 1054838 (#14 of 14): Resource leak (RESOURCE_LEAK)
leaked_storage: Variable "rop" going out of scope leaks the
storage it points to.
CID 1054836 (#5 of 5): Resource leak (RESOURCE_LEAK)
leaked_storage: Variable "rop" going out of scope leaks the
storage it points to.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
CID 1054840 (#3 of 3): Resource leak (RESOURCE_LEAK)
leaked_storage: Variable "rop" going out of scope leaks
the storage it points to.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
Close some memory leaks caused by calling new() without
freeing the memory already assigned to 'rop'.
CID 1054848 (#1-6 of 6): Resource leak (RESOURCE_LEAK)
overwrite_var: Overwriting "rop" in "rop = new_rop()"
leaks the storage that "rop" points to.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
This avoid creating a wide window between when ceph-osd is started and
when a request arrives needing a class and it is loaded. In particular,
upgrading the packages in that window may cause linkage errors (if the
class API has changed, for example).
Fixes: #5752
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
We were passing the entire parsed_args Namespace just to look at
and tack on 'format', and that's kinda silly; do it in the callers
instead.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Bug introduced in f3f92fe210.
Fixes: #5766
Backport: cuttlefish
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
We need to be able to condrestart all the ceph services on a
machine, so that we don't restart daemons that are supposed to be
stopped (e.g. broken disks).
Signed-off-by: Dan van der Ster <daniel.vanderster@cern.ch>
If not given, default to json-pretty; if given but not equal to one
of the formatter choices, return NULL as before. Remove defaulting
code in admin_socket.cc in favor of this.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
All call() routines get a format parameter; all places where
JSONFormatter was created get a new_formatter() instead.
'plain' formatting is unsupported, and help is forced to be
'json-pretty' as it was.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
requests 0.12.1 handles queryparams in the URL with embedded
spaces; requests 0.8.2 does not. Avoid the issue by quoting the
URL into expect().
Signed-off-by: Dan Mick <dan.mick@inktank.com>
We may have holes in stored maps during init_splits_between
and advance_pg. In either case, we should simply skip the
missing maps.
Fixes: #5677
Signed-off-by: Samuel Just <sam.just@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
When a nonformatted request is made, currently the only text in the
response is the (probably empty) response buffer. Add the statusmsg
as well, where the error is likely to be explained. This lets
the http client get a clue what happened.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
If the leader has and older lc than we do, and we are sharing states to
bring them up to date, we still want to also share our uncommitted value.
This particular case was broken by b26b7f6e, which was only contemplating
the case where the leader was ahead of us or at the same point as us, but
not the case where the leader was behind. Note that the call to
share_state() a few lines up will bring them fully up to date, so
after they receive and store_state() for this message they will be at the
same lc as we are.
Fixes: #5750
Backport: cuttlefish
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Joao Eduardo Luis <joao.luis@inktank.com>
To support this, we add an optional out argument to
RGWMetadatManager::put() and fill in the read_version. When the
function returns, that contains whatever the current on-disk version
of the object is (either what already existed or what we just wrote).
Signed-off-by: Greg Farnum <greg@inktank.com>
Add new STATUS_APPLIED, then specify the RGWX_UPDATE_STATUS header
based on that return code when doing metadata puts.
Add a send_response() function to RGWOp_Metadata_Put in order to
support sending back our new headers. Move the translation from
STATUS_NO_APPLY from set_req_state_err() to this function, so we
can turn different sync results into failures if necessary elsewhere.
Signed-off-by: Greg Farnum <greg@inktank.com>
Specify the param "sync-type" as one of "always", "update-by-version",
"update-by-timestamp". It defaults to always.
Signed-off-by: Greg Farnum <greg@inktank.com>
We want to be able to conditionally apply new updates:
1) if we already have a newer version than the sync is applying for some
reason (replay of logs?), we don't want to go back in time.
2) If both zones were active at the same time, then we'd like to be
able to do a merge based on timestamps.
In order to support this, we add a sync_type flag to the implementations of
RGWMetadataHandler::put, and then check the version or the mtime of the
incoming put to what we have on disk, and refuse the update if needed.
We return the 204 NoContent success code when refusing sync; for the
moment the conversion is automatic but we're going to pull it out in
the next couple commits.
This commit does not complete the feature: we don't provide an interface
for specifying a different sync protocol.
Signed-off-by: Greg Farnum <greg@inktank.com>
