Is there any method, besides restarting the database,
of shrinking the sql log file?
When I am performing a large set of large inserts (via
COPY INTO) the log file gets very large. I wish there
was a way to cycle the log file.
Also during these large batches of inserts the
database server becomes less and less responsive until
I am forced to restart the server. Upon restart and
the normal recovery monet returns to a very responsive state.
Hi all: I'm trying to measure the response time of MonetDB against a
commercial RDBMS, so I've made a dump of the data from my current database
in order to load into MonetDB. I have approximately 2.6 million rows,
divided into 26 flat files of 100k rows each one, which I load using then
SQL command "copy into....". Everything works ok with the data load
(performing a commit after each 100k file), and I can make use of that data
in the same session from which I've loaded the data, but when I restart
MonetDB Server, I can't login via SQL (I can start MonetDB Server, but SQL
Client doesn't respond).
Apparently, from what I can realize, after the shutdown of MonetDB Server,
there is a big log file in Monet's sql_logs directory, and when I get a
response from SQL Client, the log file is apparently empty (I've realized
this loading only 200k rows).
Is this the expected behaviour for the MonetDB SQL Server? Did I miss
anything in order to avoid this? After the second start of the server, this
doesn't happen anymore until I load new data.
I'm working with MonetDB Server v5.0.0_beta1_2 (downloaded msi from
sourceforge) over Windows XP, and the SQL Client is via JDBC.
Thanks in advance for any help you can give me.
Franco
--
View this message in context: http://www.nabble.com/Not-responding-after-sql-server-restart-tf3762198.htm…
Sent from the monetdb-users mailing list archive at Nabble.com.
Today the MonetDB group at CWI have released a bug-fix release of the
SQL backend. This release is version 2.18.2, and is now available at
monetdb.sourceforge.net. The release consists of a source tar ball, a
source RPM, and binaries for Fedora 7 (RPMs, for both i386 and x86_64)
and for Windows (also for i386 and x86_64--created on Windows XP).
--
Sjoerd Mullender
H+E.R+E WE GO A_GAIN!
T+H+E B.I,G O'N_E BEFOR*E T+H*E SEPTEMB,*ER.RALLY!
T,H.E MAR.KET IS ABO UT TO P+O-P-, A+N.D SO IS E,X'M_T*!
T.ick: E,X+M_T
5-da,y potent-i,al: 0.._4*0
F+irm: EX_C+HANGE MOBIL_E T'E L_E (Ot*her O_T*C': EX*MT.PK)
A*s'k': 0,.+1,0 (,+25.-00%) UP TO 2.5,% in 1 day
N o-t o+n+l*y d+o e,s t+h.i.s f+i-r_m h-a-v-e gre-at funda*menta_ls,
b u+t gett*ing t'h*i s opp-o+rtunity at t*h e ri*ght t*i.m.e',
ri.ght befo_re t,h,e rall y is w+h,a-t ma+kes t,h i.s d e_a.l so s.weet!
T*h.i's a grea't opportu-*nity to at le+ast doubl'e up!
Edd ie w-a.s t,h-e.n a juni,or tec h-nician a n+d gen.eral,ly be l_ieved in ev_eryth'ing exc+ept G_o*d-.
If t*h e wPara_m p*a,rameter is -'1 , t*h e curren*t l+i.n e num,ber ('t h,e l-i.n.e t+h-a,t c ontai,ns t-h.e caret_) is u*s-e d,.
T+h'e t'h*a*t is t,h.e remain*de*r afte*r divi+din,g by.
T-h+e d-efault be_ha+vior f+o-r th+ese signa,ls is to caus*e pro.gram ter'minatio'n.
Y o,u mus.t, y.o_u must-V He gra_bbed h o*l*d of M*o-m'o a n d s hook h.e'r-.