[Info-Ingres] hash partiton, seconday indicies and SIGSEGV...Urgent!

Hi All,



II 9.0.4 (a64.lnx/105)NPTL + p12707



As a matter of some urgency....



Over the weekend I altered a table which has several long varchar
columns to be hash partitioned 10 ways. The hash key was the same as the
btree unique key of the table.



modify gobz to btree unique on

trid, cid

with

page_size = 8192,

partition = (hash on trid, cid

partition gobz_pp1 with location=(ii_database_6),

partition gobz_pp2 with location=(ii_database_7),

partition gobz_pp3 with location=(ii_database_8),

partition gobz_pp4 with location=(ii_database_9),

partition gobz_pp5 with location=(ii_database_10),

partition gobz_pp6 with location=(ii_database_6),

partition gobz_pp7 with location=(ii_database_7),

partition gobz_pp8 with location=(ii_database_8),

partition gobz_pp9 with location=(ii_database_9),

partition gobz_pp10 with location=(ii_database_10)

)



The table has two persistent indexes named gobz_gtyp_status,
gobz_pid_gtyp.



Since the modify, any attempt to insert into the table will crash the
server with a SIGSEGV.

bb4.ctsu::[38876 , aae40200]: Mon Jun 29 09:53:55 2009
Segmentation Violation (SIGSEGV) @PC 00002b9e6a1fc4b2

RSP 0000000040895aa0 RBP 000000000000000d RSI 00002b9e8cca1288

RDI 00002b9e8cca1288 RAX 0000000000000000 RBX 00002b9e8cf5c198

RCX 00002b9e8cc9ff10 RDX 0000000000000000



-----------BEGIN STACK TRACE------------

Exception in stack unwind

Trying again without symbols

Exception in stack unwind

-----------END STACK TRACE----------

bb4_ctsu::[38876 , 00002aaaaae40200]: Mon Jun 29 09:53:55
2009 E_SC0221_SERVER_ERROR_MAX Error count for server has been
exceeded.

bb4_ctsu::[38876 , 00002aaaaae40200]: Query: EXECUTE e0
using ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V ,
~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V , ~V ^@

bb4_ctsu::[38876 , 00002aaaaae40200]: LQuery: PREPARE e0
FROM INSERT INTO gobz
(cid,trid,xu_hid,aid,rid,xu_ts,xu_at,xu_ca,chain,a uthor,tsig,pver,xu_chk
,xu_flag,pid,gtyp,xb_outr,xb_innr,status,xu_ovalid ,xub_osr,xu_ivalid,xub
_isr) VALUES ( ? , ? , ? , ? , ? , ? , ? , ? , ? , ? , ? , ? , ? , ? , ?
, ? , ? , ? , ? , ? , ? , ? , ? )^@



Sometimes the log shows...

bb4_ctsu::[49132 , 00002aaaaba12200]: Mon Jun 29 09:25:05
2009 E_DM9C02_MISSING_TCB_INFO TCB information was missing. Page
12584 in table 243,-2147477243 of database ace_pack_live did not have
the required Table Control Block information for its secondary indexes.

bb4_ctsu::[49132 , 00002aaaaba12200]: Mon Jun 29 09:25:05
2009 E_DM9C03_INDEX_TCB_INFO TCB information exists for secondary index
gobz_gtyp_status (243,6456), in database ace_pack_live.

bb4_ctsu::[49132 , 00002aaaaba12200]: Mon Jun 29 09:25:05
2009 E_DM9C03_INDEX_TCB_INFO TCB information exists for secondary index
gobz_pid_gtyp (243,6455), in database ace_pack_live.

bb4_ctsu::[49132 , 00002aaaaba12200]: Mon Jun 29 09:25:05
2009 E_DM9449_DMD_CHECK DMD_CHECK called from file
/devsrc/ingresr30/b105/src/back/dmf/dmp/dm0p.c line 14299.

Exception in stack unwind

bb4_ctsu::[49132 , 00002aaaaba12200]: Mon Jun 29 09:25:05
2009 E_DM9300_DM0P_CLOSE_PAGE Buffer still fixed.



I've tried dropping and recreateing the secondary indicies...No good.

I've tried just dropping the secondary indicies...No good.

I'm now going to rebuild the table with nopartitions and see if that
solves the problem.



The rebuild will take about 3-4 hours...In the mean time does anyone
have a suggestion as to what might be causing the problem.



Martin Bowes