2112945 - Error 814 Keep count of buffer xxx in cache 'cachename' holding logical page nnn in database 'db' has become negative - SAP ASE

• In some rare circumstance the ASE server can encounter Error 814 with a stack trace, for example:

Error: 814, Severity: 20, State: 7
Keep count of buffer '0x11462308c18' in cache 'tempdb_cache' holding logical page '651777' in database 'My_DB' has become negative.

• This error is not triggered by a specific command, but it seems to be most common when executing the update ... statistics command, for example:

 kernel  SQL causing error: update index statistics TABLE_NAME with sampling = 1 percent
 server  SQL Text: update index statistics TABLE_NAME with sampling = 1 percent

• Stacktrace includes the following modules:

bufunkeep_cache
bufallunkeep
closetable
ups_end_statglobalnosort
ups_calctabdata
ups__updstats
ups__updstats_driver
update_statistics

• Then, the ASE server could encounter other kinds of subsequent errors, for example:

Error: 837, Severity: 21, State: 1
Attempted to grab wash header

Error: 839, Severity: 21, State: 1  Grabbed buffer dirty but not writing : BUF=0x0000011462308c18, MASS=0x0000011462308c18, Buf#=0, page=0x0000011443760000, dbid=15, Mass vdevno=34, vpage=2450434, Buf lpage=1225217, Mass stat=0x4011000, Buf stat=0x1, size=4096, cid=1 ('tempdb1_cache'), Pinned xdes=0x0000000000000000, spid=0.

    NOTE:  Error 839 can occur a huge number of times after Error 814, but always on the same Buffer Number 0x11462308c18.

SAP Adaptive Server Enterprise (ASE) 15.7

750807, CR750807, CR#750807, _$o1cexlA0.bufunkeep_cache, _$o1cexpL0.bufallunkeep, _$o1cexpM0.closetable, keepcount, spinlock, DES, unkeep , KBA , BC-SYB-ASE , Sybase ASE Database Platform (non Business Suite) , Bug Filed