2458106 - Log Segment Full Caused by XFS Blocksize and ASE Server Page Size Mismatch

If XFS file system is being used for ASE server on Linux,

• The log segment can not get cleared fast enough even when the database option 'trunc log on chkpt' is turned on
• In the output of 'sp_sysmon', you can see 'Avg Time per Normal Chkpt' is high, for example,
  
  Avg Time per Normal Chkpt      34.90304 seconds
  
  
• If 'disk mirroring' is enabled, the 'Mirror Semaphore Waited' for some device is very high
  
    Device:                                                                      
      /mnt/MULMP1_15/wwsdpp/wwsdpp_data11                                        
      wwsdpp_data11                 per sec      per xact       count  % of total
    -------------------------  ------------  ------------  ----------  ----------
      Reads                                                                      
        APF                             0.0           0.0           0       0.0 %
        Non-APF                        50.5           0.5      178807       6.4 %
      Writes                          743.4           6.9     2631685      93.6 %
    -------------------------  ------------  ------------  ----------  ----------
    Total I/Os                        793.9           7.3     2810492      23.2 %
  
    Mirror Semaphore Granted          316.8           2.9     1121385      25.9 %
    Mirror Semaphore Waited           907.1           8.4     3211190      74.1 %

• SAP Adaptive Server Enterprise (ASE) 15.7
• SAP Adaptive Server Enterprise (ASE) 16.0
• Linux on Power 64bit
• Linux on x86_64 64bit

checkpoint, trunc log on chkpt, XFS filesystem, ASE pagesize, Avg Time per Normal Chkpt, Mirror Semaphore Waited , KBA , BC-SYB-ASE , Sybase ASE Database Platform (non Business Suite) , BC-DB-SYB , Business Suite on Adaptive Server Enterprise , Problem