Hi,
2004-12-28 kl. 07.29 skrev NiCK:
> Hi,expert,
>
> 3 databases , 31 tables , 3,000,000 records
Some additional info is needed
1) Number of indexes on tables (Most likely default is ok, but a check
is needed)
2) Size of primary keys (affects how much IndexMemory is needed)
3) Size of records in tables, mean record size is ok (affects how much
DataMemory is needed)
If you are doing large transactions you also need to know the maximum
amount of records involved
in a transaction.
Some kind of hint on if you are doing high update load is also
beneficial.
Then some calculations can be applied to figure out most of the
appropriate configuration parameters.
Rgrds Mikael
PS: The config file you supplied in a previous email looks fairly ok
given that your machines are equipped
with 2 GB of memory. The problems you got looked like bugs, so
supplying the important parts of the tracefile
for the crash can usually give a pointer to the problem and sometimes
even a pointer to the solution.
The important parts of the trace log is around where the line number
logs end and the signal logs start.
DBLQH 3045 here means that line 3045 in DblqhMain.cpp was passed. So in
the example below the
crash occurred while executing ATTRINFO (sent to Dblqh) in Dbtup on
line 442 (Dbtup is a block with many
files and the translation from 32292 to line and file is provided by a
list in Dbtup.hpp, same with Dbtux).
DBTC 002070 002104 001984 002149 003296 003342
DBLQH 003200 003372 016091
DBTUP 000162
DBLQH 003509 003539
DBLQH 002900 003068 002920
DBTUP 000162 000136
DBLQH 003045
DBACC 001630 004547 004585 004597
DBLQH 003774
DBTUP 000661 000075 000075 014127 014127 014127 014127 014127
014127 014127 014131 032285 032292 000442
--------------- Signal ----------------
r.bn: 247 "DBLQH", r.proc: 2, r.sigId: 268393 gsn: 4 "ATTRINFO" prio: 1
s.bn: 245 "DBTC", s.proc: 2, s.sigId: 268391 length: 4 trace: 1 #sec: 0
fragInf: 0
H'00008a38 H'000004bf H'00200300 H'00050000
--------------- Signal ----------------r.bn: 247 "DBLQH", r.proc: 2,
r.sigId: 268392 gsn: 316 "LQHKEYREQ" prio: 1s.bn: 245 "DBTC", s.proc:
2, s.sigId: 268391 length: 19 trace: 1 #sec: 0 fragInf: 0
ClientPtr = H'00008a38 hashValue = H'27adbca7 tcBlockRef = H'00f50002
transId1 = H'000004bf transId2 = H'00200300 savePointId = H'00000000
Op: 0 Lock: 1 Flags: ScanInfo/noFiredTriggers: H'6a
> 2 intel base servers
> how organize is the best? and how to configure config.ini?
> Who have experience for this environment?
>
>
>
>
>
> --
> MySQL Cluster Mailing List
> For list archives: http://lists.mysql.com/cluster
> To unsubscribe:
> http://lists.mysql.com/cluster?unsub=1
>
>
Mikael Ronström, Senior Software Architect
MySQL AB, www.mysql.com
Clustering:
http://www.infoworld.com/article/04/04/14/HNmysqlcluster_1.html
http://www.eweek.com/article2/0,1759,1567546,00.asp
