Oracle RAM shared memory errors for UNIX Oracle Database Tips by Donald Burleson

Oracle MOSC Note 15566.1 Unix Semaphores and Shared Memory Explained, has a great summary of Oracle error codes relating to server memory errors:

ORA-7306, ORA-7336, ORA-7329

  Oracle received a system error on a shmget() call.  The system error

  should be reported.  There are a few possibilities:

  1) There is insufficient shared memory available.  This is

     indicated by the operating system error ENOSPC.  Most likely, SHMMNI

     is too small.  Alternatively, there may shared memory already

     allocated; if it is not attached, perhaps it can be freed.  Maybe

     shared memory isn't configured in the kernel.

  2) There is insufficient memory available.  Remember, shared memory

     needs pages of virtual memory.  The system error ENOMEM indicates there

     is insufficient virtual memory.  Swap needs to be increased, either by

     adding more or by freeing currently used swap (i.e. free other shared

     memory, kill other processes)

  3) The size of the shared memory segment requested is invalid.  In this

     case, EINVAL is returned by the system. This should be very rare - however, 

     it is possible.  This can occur if SHMMAX is not a mulitple of page

     size and Oracle is trying a multi-segment model.  Remember that Oracle

     rounds its calculation of SHMMAX to a page boundary, so it may have

     rounded it up past the real SHMMAX! (Whether this is a bug is

     debatable).

  4) The shared memory segment does not exist.  This would be indicated

     by the system error ENOENT.  This would never happen on startup; it

     only would happen on connects.  The shared memory most likely has been

     removed unexpectedly by someone or the instance is down.

  ORA-7307, ORA-7337, ORA-7320

  Oracle received a system error on a shmat() call.  The system should be

  reported.  There a a few possibilities:

  1) The attach address is bad.  If this is the cause, EINVAL is returned

     by the system.  Refer to the section on the attach address to see why

     the attach address might be bad.  This may happen after enlarging the

     SGA.

  2) The permissions on the segment do not allow the process to attach

     it.  The operating system error will be EACCES.  Generally the cause of

     this is either the setuid bit is not turned on for the oracle

     executable, or root started the database (and happens to own the shared

     memory).  Normally, this would be seen only on connects.

  3) The process cannot attach any more shared memory segments.  This

     would be accompanieed by the system error EMFILE.  SHMSEG is too

     small.  Note that as long as SHMSEG is greater than SS_SEG_MAX, you

     should never see this happen.

  ORA-7329, ORA-7334

  Oracle has determined the SGA needs too many shared memory segments.  Since you

  can't change the limit on the number of segments, you should instead increase

  SHMMAX so that fewer segments are required.

  ORA-7339

  Oracle has determined it needs too many semaphore sets.  Since you

  can't change the limit on the number of semaphore sets, you should

  increase SEMMSL so fewer sets are required.

  ORA-7250, ORA-7279, ORA-7252, ORA-27146

  Oracle received a system error on a semget() call.  The system error should be

  reported.  There should be only one system error ever returned with

  this, ENOSPC.  This can mean one of two things.  Either the system

  limit on sempahore sets has been reached or the system limit on the

  total number of semaphores has been reached.  Raise SEMMNI or SEMMNS,

  as is appropriate, or perhaps there are some semaphore sets which can

  be released.  In the case of ORA-7250, ORANSEMS may be set too high

  (>SEMMSL).  If it is, raise SEMMSL or decrease ORANSEMS.

  ORA-7251

  Oracle failed to allocate even a semaphore set of only one semaphore.  It is

  likely that semaphores are not configured in the kernel.

  ORA-7318

  Oracle could not open the sgadef file.  The system error number will be

  returned.  There are a few possible causes:

  1) The file doesn't exist.  In this case, the system error ENOENT is

     returned.  Maybe ORACLE_SID or ORACLE_HOME is set wrong so that Oracle

     is looking in the wrong place.  Possibly the file does not exist (in this 

     case, a restart is necessary to allow connections again).

  2) The file can't be accessed for reading.  The operating system error returned

     with this is EACCES.  The permissions on the file (or maybe

     directories) don't allow an open for reading of the sgadef file.  It

     might not be owned by the oracle owner.  The setuid bit might not be

     turned on for the oracle executable.

  ORA-7319

  Oracle did not find all the data it expected when reading the

  sgadef<SID>.dbf file.  Most likely the file has been truncated. The

  only recovery is to restart the instance.

  ORA-7430

  Oracle expected a key to be used for the segment which does not match the

  key stored in the shared memory and semaphore data structure.  This probably

  indicates a corruption of the sgadef file (in version 6) or

  the data in the first segment of the SGA (in version 7).  A restart of

  the instance is probably necessary to recover in that case.  It may

  also be a key collision problem and Oracle is attached to the wrong

  instance.

  ORA-7321

  Oracle was able to attach the segment, but not at the address it

  requested.  In most cases, this would be caused by corrupted data in

  the sgadef file (in version 6) or the first segment of the SGA (in

  version 7).  A restart of the database may be necessary to recover.

  ORA-7324, ORA-7325

  Oracle was unable to allocate memory.  Most likely, the heap (data

  segment) has grown into the bottom of the SGA.  Relocating the SGA to a

  higher attach address may help, but there may be other causes.  Memory

  leaks can cause this error.  The init.ora parameter sort_area_size may be

  too large, decreasing it may resolve the error.  The init.ora parameter

  context_incr may also be too large, decreasing it may resolve this

  ORA-7264, ORA-7265

  Oracle was unable to decrement/increment a semaphore.  This generally

  is accompanied by the system error EINVAL and a number which is the

  identifier of the semaphore set.  This is almost always because the

  semaphore set was removed, but the shadow process was not aware of it

  (generally due to a shutdown abort or instance crash).  This error

  is usually ignorable.

If you like Oracle tuning, you may enjoy my new book "Oracle Tuning: The Definitive Reference", over 900 pages of BC's favorite tuning tips & scripts.  You can buy it direct from the publisher for 30%-off and get instant access to the code depot of Oracle tuning scripts.