This is an automatically generated e-mail. To reply, visit: http://reviewboard-openlmi.rhcloud.com/r/1760/

On May 23rd, 2014, 7:18 a.m. UTC, Michal Minar wrote:

mof/60_LMI_Journald.mof (Diff revision 1)
228
    [ Implemented(true), Description (
229
          "Denotes numerical effective user ID of the process that sent the "
230
          "message. This ID is system specific and usually maps to a local "
231
          "POSIX account." ) ]
232
    sint32 UserID;
233
234
    [ Implemented(true), Description (
235
          "Denotes numerical effective group ID of the process that sent the "
236
          "message. This ID is system specific and usually maps to a local "
237
          "POSIX account." ) ]
238
    sint32 GroupID;
239
240
    [ Implemented(true), Description (
241
          "Denotes numerical ID of the process that sent the message." ) ]
242
    uint32 ProcessID;

Having 2 IDs signed and one unsigned looks weird. If I take a look at CIM schema, UserID and GroupID properties are strings but there's also CIM_UnixProcess with RealUserID and RealGroupID. They are uint64 properties.

IMHO in this case there's no harm in sticking to CIM convention.

Hmm, right, good idea. I'm tempted to use String to stay consistent but it's really odd. Besides, numbers allow easier range filtering in WQL/CQL and type conversion is easy in Python/lmishell. Anyway, changed to uint64.


On May 23rd, 2014, 7:18 a.m. UTC, Michal Minar wrote:

mof/60_LMI_Journald.mof (Diff revision 1)
250
      Values { "kern", "user", "mail", "daemon", "auth", "syslog", "lpr",
251
               "news", "uucp", "clock", "authpriv", "ftp", "ntp", "audit",
252
               "alert", "cron", "local0", "local1", "local2", "local3",
253
               "local4", "local5", "local6", "local7" } ]

I don't know much about logging facilities, but is this information always known? I'm missing something like "unknown" in possible values.

If the journal item field is unknown or missing, the property is simply not set. I've added more conditions to ensure values are in the range.


- Tomáš


On May 23rd, 2014, 2:47 p.m. UTC, Tomáš Bžatek wrote:

Review request for OpenLMI Developers.
By Tomáš Bžatek.

Updated May 23, 2014, 2:47 p.m.

Repository: openlmi-providers

Description

This is useful e.g. for indication filtering.

Diffs

  • mof/60_LMI_Journald.mof (938b3d282029d7b8c91d6ccc03efab150a79c22d)
  • src/journald/instutil.c (b3987bccf2b4a641e384ef96f3de260c58a5ac44)

View Diff