On 2/17/2017 2:59 PM, Anders Widell wrote:
> we should check that our README refers to the RFC somewhere.
That will be even good for completely understand the New MDS log format.
-AVM
On 2/17/2017 2:59 PM, Anders Widell wrote:
> Hi!
>
>
> On 02/17/2017 10:16 AM, A V Mahesh wrote:
>> Hi Anders Widell,
>>
>> ACK with following ( tested as well )
>>
>> On 2/17/2017 2:22 PM, Anders Widell wrote:
>>> Yes the max value is INT_MAX (2147483647), after which the next
>>> sequence number will start from 1 again.
>>
>> Do we have to mention the things in some README ?
>
> This limit is documented in the RFC, but we should check that our
> README refers to the RFC somewhere.
>
>>
>> -AVM
>>
>> On 2/17/2017 2:22 PM, Anders Widell wrote:
>>> See below.
>>>
>>> regards,
>>>
>>> Anders Widell
>>>
>>>
>>> On 02/17/2017 06:21 AM, A V Mahesh wrote:
>>>> Hi Anders Widell,
>>>>
>>>> Can you please clarify below :
>>>>
>>>> 1) is this meta sequenceId is rotates after reaching some value ?
>>>> ( Max value )
>>>
>>> Yes the max value is INT_MAX (2147483647), after which the next
>>> sequence number will start from 1 again.
>>>
>>>>
>>>> 2) is this meta sequenceId persistent even after application
>>>> respawn ?
>>>
>>> No, but you can detect application restart using the PID, which is
>>> also included in the log message.
>>>
>>>>
>>>> =============================================================================================================================================================
>>>>
>>>>
>>>>
>>>> <143>1 2017-02-17T10:44:21.59772+05:30 PL-1 osafimmloadd 10183 -
>>>> [meta sequenceId="431572"] MDS_SND_RCV : calling cb ptr enc or enc
>>>> flatin mcm_msg_encode_full_or_flat_and_send
>>>> <143>1 2017-02-17T10:44:21.597733+05:30 PL-1 osafimmloadd 10183 -
>>>> [meta sequenceId="431573"] >> mds_get_subtn_res_tbl_by_adest
>>>>
>>>> <143>1 2017-02-17T10:44:21.597771+05:30 PL-1 osafimmloadd 10183 -
>>>> [meta sequenceId="431576"] << mcm_msg_encode_full_or_flat_and_send
>>>> <142>1 2017-02-17T10:44:21.597784+05:30 PL-1 osafimmloadd 10183 -
>>>> [meta sequenceId="431577"] MDTM: User Sending Data lenght=384 From
>>>> svc_id = IMMA_OM(26) to svc_id = IMMND(25)
>>>> <143>1 2017-02-17T10:44:21.597798+05:30 PL-1 osafimmloadd 10183 -
>>>> [meta sequenceId="431578"] MDTM: Sending message with Service
>>>> Seqno=7482, TO Dest_id=<0x0002020f:9253>
>>>>
>>>> <143>1 2017-02-17T10:44:21.597875+05:30 PL-1 osafimmnd 9253 - [meta
>>>> sequenceId="1532911"] >> mds_vdest_tbl_get_role
>>>> <143>1 2017-02-17T10:44:21.597879+05:30 PL-1 osafimmnd 9253 - [meta
>>>> sequenceId="1532912"] << mds_vdest_tbl_get_role
>>>>
>>>> =============================================================================================================================================================
>>>>
>>>>
>>>>
>>>> -AVM
>>>>
>>>>
>>>> On 2/13/2017 8:11 PM, Anders Widell wrote:
>>>>> src/mds/mds_log.cc | 24 ++++++++++++++++--------
>>>>> 1 files changed, 16 insertions(+), 8 deletions(-)
>>>>>
>>>>>
>>>>> Use the meta sequenceId structured element for message sequence
>>>>> numbers instead
>>>>> of the MSGID field, as per RFC-5424.
>>>>>
>>>>> Also use a mutex to protect the UnixSocket instance, since users
>>>>> of this class
>>>>> will be reponsible for thread safety after ticket [#2293].
>>>>>
>>>>> diff --git a/src/mds/mds_log.cc b/src/mds/mds_log.cc
>>>>> --- a/src/mds/mds_log.cc
>>>>> +++ b/src/mds/mds_log.cc
>>>>> @@ -27,7 +27,6 @@
>>>>> #include <sys/time.h>
>>>>> #include <time.h>
>>>>> #include <unistd.h>
>>>>> -#include <atomic>
>>>>> #include <cstdarg>
>>>>> #include <cstdio>
>>>>> #include <cstring>
>>>>> @@ -35,6 +34,7 @@
>>>>> #include "base/buffer.h"
>>>>> #include "base/log_message.h"
>>>>> #include "base/macros.h"
>>>>> +#include "base/mutex.h"
>>>>> #include "base/ncsgl_defs.h"
>>>>> #include "base/osaf_utility.h"
>>>>> #include "base/time.h"
>>>>> @@ -54,12 +54,14 @@ class MdsLog {
>>>>> uint32_t proc_id, const char* socket_name);
>>>>> void LogInternal(base::LogMessage::Severity severity, const
>>>>> char *fmt,
>>>>> va_list ap);
>>>>> + static constexpr const uint32_t kMaxSequenceId =
>>>>> uint32_t{0x7fffffff};
>>>>> static MdsLog* instance_;
>>>>> const base::LogMessage::HostName host_name_;
>>>>> const base::LogMessage::AppName app_name_;
>>>>> const base::LogMessage::ProcId proc_id_;
>>>>> - std::atomic<uint64_t> msg_id_;
>>>>> + uint32_t sequence_id_;
>>>>> base::UnixClientSocket log_socket_;
>>>>> + base::Mutex mutex_;
>>>>> DELETE_COPY_AND_MOVE_OPERATORS(MdsLog);
>>>>> };
>>>>> @@ -72,8 +74,9 @@ MdsLog::MdsLog(const char* host_name, co
>>>>> host_name_{base::LogMessage::HostName{host_name}},
>>>>> app_name_{base::LogMessage::AppName{app_name}},
>>>>> proc_id_{base::LogMessage::ProcId{std::to_string(proc_id)}},
>>>>> - msg_id_{0},
>>>>> - log_socket_{socket_name} {
>>>>> + sequence_id_{1},
>>>>> + log_socket_{socket_name},
>>>>> + mutex_{} {
>>>>> }
>>>>> /*****************************************************
>>>>> @@ -130,16 +133,21 @@ void MdsLog::Log(base::LogMessage::Sever
>>>>> void MdsLog::LogInternal(base::LogMessage::Severity severity,
>>>>> const char *fmt,
>>>>> va_list ap) {
>>>>> - uint64_t id = msg_id_++;
>>>>> + base::Lock lock(mutex_);
>>>>> + uint32_t id = sequence_id_;
>>>>> + sequence_id_ = id < kMaxSequenceId ? id + 1 : 1;
>>>>> base::Buffer<256> buffer;
>>>>> - base::LogMessage::Write(base::LogMessage::Facility::kLocal0,
>>>>> + base::LogMessage::Write(base::LogMessage::Facility::kLocal1,
>>>>> severity,
>>>>> base::ReadRealtimeClock(),
>>>>> host_name_,
>>>>> app_name_,
>>>>> proc_id_,
>>>>> - base::LogMessage::MsgId{std::to_string(id)},
>>>>> - {},
>>>>> + base::LogMessage::MsgId{""},
>>>>> + {{base::LogMessage::SdName{"meta"},
>>>>> + {base::LogMessage::Parameter{
>>>>> + base::LogMessage::SdName{"sequenceId"},
>>>>> + std::to_string(id)}}}},
>>>>> fmt,
>>>>> ap,
>>>>> &buffer);
>>>>
>>>
>>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Opensaf-devel mailing list
Opensaf-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/opensaf-devel
