| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Invalid lengths come in two flavours: ones that correctly point at the
end of an AVP's payload but don't agree with its type, and ones that
point elsewhere. The former are relatively harmless but the latter leave
no way to recover AVP boundaries, which typically results in failure to
decode subsequent AVP's in the message in question.
In the case that AVP Length points past the end of the message, diameter
incorrectly regarded the error as 5009, INVALID_AVP_BITS: not right
since the error has nothing to do with AVP Flags. Ditto if the length
was less than 8, a minimal header length. Only in the remaining case was
the detected error 5014, INVALID_AVP_LENGTH. However, in this case it
slavishly followed RFC 3588 in suggesting the undecodable AVP as
Failed-AVP, thereby passing the woeful payload back to the sender to
have equal difficulty decoding. Now follow RFC 6733 and suggest an AVP
with a zero-filled payload.
|
|
When setting Failed-AVP in a message record, it was never tested that
the field was actually present. RFC 6733 says it should be, 3588 says
MAY.
|
|
To return what RFC 6733 says. 3588 says less so follow 6733, even
though the extra specification of 6733 means that it isn't strictly
backwards compatible. In particular, 6733 says to send a zero'd payload
or none at all while 3588 says to send the offending AVP, despite the
fact that the peer will likely have equal difficulty in decoding it.
The testcases now fail, which will be remedied in subsequent commits.
|
|
There are still enslave/1 failures on some hosts despite a 2 minute
timetrap.
|
|
Such a length caused decode of a message with valid (24-bit) length to
fail. Note that the error detected is wrong: it should be 5014
(INVALID_AVP_LENGTH), not 3009 (INVALID_AVP_BITS). This will be dealt
with by OTP-11007.
|
|
Creating examples/code and examples/dict in parallel can fail when
examples doesn't exists. This has been seen on FreeBSD.
|
|
Point was to test that Session-Id was not undefined. Instead, test case
send_error just returned false.
|
|
This is initially to identify the source of some flakiness on Windows.
|
|
Due to sporadic timeouts one some (slow) hosts.
|
|
In particular, have the resulting badmatch contain the starting and
ending time. There are still sporadic failures on slow hosts.
|
|
* anders/diameter/watchdog_leak/OTP-11019:
Minor doc fix
Add testcase to exercise reconnect behaviour
Fix watchdog table leak
|
|
|
|
|
|
|
|
In particular, remove timing dependence by using diameter_reg:wait/1 to
wait on the term registered by diameter_{tcp,sctp} when opening a
listening socket.
|
|
Note that the semantics of client:connect/1 have changed slightly: the
second element in an argument 3-tuple is a remote address, the local
address being the transport module's default. Previously it was
interpreted as a common local/remote address.
|
|
Use the default address address (as selected by gen_tcp) if none is
configured, passing it in the new 'connected' message introduced by the
previous commit.
The corresponding update to diameter_sctp has to wait until problems
with inet:sockname/1 are resolved: the function currently only returns
one address, and sometimes {0,0,0,0}. See OTP-11018.
|
|
A transport module can return a local address list from its start/3
function in order to specify addresses to be used as Host-IP-Address
during capabilities exchange. Now allow addresses to be communicated in
a 'connected' message in the case of a connecting transport, so that
diameter_tcp (in particular) can make local address configuration
optional, communicating the gen_tcp default after connection
establishment instead.
|
|
A service process maintains a table keyed on watchdog process pids. When
a watchdog process dies the corresponding entry should be removed but
this was broken in commit f115a9f7, causing entries with watchdog state
DOWN to accumulate.
Watchdog processes die as a result of diameter:remove_transport/2, or
when a peer reestablishes a connection in the listening case. Neither is
typically a frequent occurrence.
The fault manifests itself in the return value of
diameter:service_info(SvcName, transport), which displays entries for
watchdog processes that are no longer alive.
|
|
* anders/diameter/service_config/OTP-11017:
Fix handling of unknown options to diameter:start_service/2
|
|
Diameter = the protocol
diameter = the Erlang application
|
|
{error, Reason} is now returned, instead of the options being ignored.
Note that diameter:add_transport/2 purposely ignores unknown options and
that the behaviour is documented. This is historic: some users depend on
it in order to store their own options for identifying transport config,
instead of using the reference returned by add_transport.
|
|
* anders/diameter/app_not_configured/OTP-10972:
Fix faulty sequence validation
Add config suite
Deal with config errors detected at transport start less brutally
Move most transport_opt() validation into diameter_config
Minor doc/spec fix
Minor diameter_lib cleanup
|
|
The validation of {sequence, {H,N}} incorrectly checked that H was an
N-bit integer, instead of the intended 32-N.
|
|
To verify return values from diameter:start_service/2 and
diameter:add_transport/2 when passing various config.
|
|
Crashing watchdog and peer_fsm processes was somewhat unseemly. Emit an
error report and die silently instead.
|
|
Faulty configuration was previously passed directly on to watchdog and
peer_fsm processes, diameter:add_transport/2 happily returning ok and
the error resulting on failure of watchdog and/or peer_fsm processes.
Now check for errors before getting this far, returning {error, Reason}
from diameter:add_transport/2 when one is detected. There are still
some errors that can only be detected after transport start (eg. a
misbehaving callback) but most will be caught early.
|
|
'infinity' is a valid transport_config timeout.
|
|
Remove unused functions, add dialyzer specs, make wait/1 less fallible.
|
|
|
|
Instead of from the installation.
|
|
Fix errors and omissions related to dictionary compilation.
|
|
From compiler suite.
|
|
That is, code installed under examples/code in an installation.
|
|
Fix a broken include in example code, remove an inappropriate ct:pal/2
outside of a testcase, echo more info from test/Makefile.
|
|
|
|
|
|
|
|
|
|
RFC 6733 has changed the arity of Vendor-Id in this Grouped AVP, from 1*
in RFC 3588 to 1 in RFC 6773. This impacts the generated dictionary
modules: Vendor-Id is expected to be list-valued in the 3588 dictionary,
integer-valued in the 6733 dictionary. This, in turn, breaks the
independence of capabilities configuration on a service or transport
from the dictionary that will be used to encode an outgoing CER or CEA.
This commit fixes this by massaging any Vendor-Specific-Application-Id
config as appropriate when constructing CER or CEA for a given
dictionary.
|
|
|
|
|
|
Allow both share_peers and use_shared_peers to be a list of nodes, or a
function that returns a list of nodes.
|
|
|
|
|
|
This is the functionality that allows transports to be shared between
identically-named services on different nodes, which has been neither
documented nor tested (until now).
|
|
Make it just a number of timeouts, without a new DWR being sent.
|
|
|
|
Count watchdog messages instead when this is an alternative.
|
