- users-list - monetdb.org

Query takes too much time
by Ivan Videla 05 Dec '12

05 Dec '12

Hi All, I have a table with 268 millions of records from supermarket transactions tran_id, prod_id, quantity, price, cost and i want to make a calculation like sum(quantity*(price-cost)). But this query takes too much time. I want to know if i can improve the results using an index. Or monetdb itself use an index on-the-fly. If not i would like to know how can i make an index to improve the execution time. Regards, Ivan _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

3 12

Failed to load 128GB CSV file using Oct2012
by Ying Zhang 05 Dec '12

05 Dec '12

Hai all, I try to load a CVS file of ~128GB in the ssdb branch (based on Oct2012), but after quite sometime, it failed with: #GDKrealloc(19598973840) fails, try to free up space [memory in use=22695528216,virtual memory in use=138875848472] #GDKrealloc(19598973840) result [mem=22695528216,vm=138875848472] In the last Sep., I was able to execute the same query with the same data. But now it takes much longer and ends in a crash. For INS1-ers, the command to start the server: /export/scratch2/zhang/monet-install/ssdb/debug/bin/mserver5 --set gdk_dbfarm=/export/scratch2/zhang/dbfarm --dbname=ssdb --set mapi_usock=/export/scratch2/zhang/dbfarm/ssdb/.mapi.sock --set monet_vault_key=/export/scratch2/zhang/dbfarm/ssdb/.vaultkey --set gdk_nr_threads=8 --set max_clients=64 --set sql_optimizer=default_pipe --set mapi_port=60000 The command to start the query: cd /ufs/zhang/papers/ssdb/monetdb ./run.sh loadimg small output Does anyone have an idea? If anymore information is needed, please let me know. I'm re-running the query to get a GDB stack trace. Thanks! Jennie _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

1 0

INET query/type conversion issues
by Percy Wegmann 04 Dec '12

04 Dec '12

Hi MonetDB Team, I encountered a strange problem related to querying against an INET column. To reproduce: percy@percy-sandbox:~$ uname -a Linux percy-sandbox 3.2.0-33-generic #52-*Ubuntu* SMP Thu Oct 18 16:29:15 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux percy@percy-sandbox:~$ mclient -u monetdb -d click password: Welcome to mclient, the MonetDB/SQL interactive terminal (Oct2012) Database: *MonetDB v11.13.3 (Oct2012)*, 'mapi:monetdb://percy-sandbox:50000/click' Type \q to quit, \? for a list of available commands auto commit mode: on sql>create table test_table(col inet); operation successful (27.394ms) sql>insert into test_table values(*'127.0.0.1'*); 1 affected row (20.370ms) sql>insert into test_table values(*inet '127.0.0.1'*); 1 affected row (19.654ms) sql>insert into test_table values(*cast('127.0.0.1' as inet)*); 1 affected row (25.038ms) sql>select count(*) from test_table where col = '127.0.0.1'; +------+ | L1 | +======+ | *3* | +------+ 1 tuple (1.547ms) sql>select count(*) from test_table where col = inet '127.0.0.1'; +------+ | L1 | +======+ | *0* | +------+ 1 tuple (0.946ms) sql>select count(*) from test_table where col = cast('127.0.0.1' as inet); +------+ | L1 | +======+ | *2* | +------+ 1 tuple (0.982ms) sql>select count(*) from test_table where col = inet '127.0.0.1'; +------+ | L1 | +======+ | *2* | +------+ 1 tuple (0.937ms) sql>select count(*) from test_table where col = inet '127.0.0.1'; +------+ | L1 | +======+ | *0* | +------+ 1 tuple (0.801ms) Expected Result: All queries return a count of 3 Actual Result: Queries return different counts depending on how INET value was bound to query. Particularly interesting is that the output when using the inet '127.0.0.1' syntax is non-deterministic. Does this ring any bells? Should I log it? Cheers, -- Percy Wegmann +1 512 637 8500 ext 148 _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

1 1

Manually Loading all SQL Script
by swapnil joshi 04 Dec '12

04 Dec '12

Hi , My SQL Script is loaded only one's, When i am creating new database using following command I have copy and paste my terminal command swapnil@swapnil-lp:~$ sudo monetdb create test5 [sudo] password for swapnil: Sorry, try again. [sudo] password for swapnil: created database in maintenance mode: test5 swapnil@swapnil-lp:~$ sudo monetdb release test5 taken database out of maintenance mode: test5 swapnil@swapnil-lp:~$ sudo monetdb start test5 starting database 'test5'... done swapnil@swapnil-lp:~$ mclient test5 Welcome to mclient, the MonetDB/SQL interactive terminal (Jul2012-SP1) Database: MonetDB v11.11.7 (Jul2012-SP1), 'mapi:monetdb://swapnil-lp:50000/test5' Type \q to quit, \? for a list of available commands auto commit mode: on sql> After that if i am checking merovingian.log (Monetdb log file) 2012-12-04 17:57:20 MSG test2[11694]: # SQL catalog created, loading sql scripts once 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 09_like.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 10_math.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 11_times.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 12_url.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 13_date.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 14_inet.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 15_history.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 16_tracelog.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 17_compress.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 18_dictionary.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 19_cluster.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 20_vacuum.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 21_dependency_functions.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 22_clients.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 23_skyserver.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 24_zorder.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 25_debug.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 40_geom.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 80_udf.sql 2012-12-04 17:57:20 MSG test2[11694]: # loading sql script: 99_system.sql How will i manually Lode this script because if i have change in any *.sql file each time I need to create new database How will i avoid this manual work? Please help me!!!! :) -- Regards, Swapnil K. Joshi _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

2 1

Can Monetdb drop Calculation of dropped column?
by swapnil joshi 04 Dec '12

04 Dec '12

Hi I have executed following SQL Statement : *sql>select product_id, sum(unit_sales),avg(unit_sales), max(unit_sales), min(unit_sales),sum(store_sales),avg(store_sales), max(store_sales), min(store_sales), sum(store_cost),avg(store_cost), max(store_cost), min(store_cost) from sales_fact_1997 group by product_id; It Gives following result: (In my table very large data there for i have given starting and ending rows of result set) +-----------+--------------+--------------------------+--------------+--------------+--------------+--------------------------+--------------+--------------+ | product_i | L1 | L2 | L3 | L4 | L5 | L6 | L7 | L10 |> : d : : : : : : : : :> +===========+==============+==========================+==============+==============+==============+==========================+==============+==============+ | 337 | 333644.7415 | 3.4849771407382648 | 6.9998 | 0.0001 | 1145678.6601 | 11.966812134157813 | 23.9986 | 0.0005 | | 1512 | 336475.1207 | 3.4816655356884185 | 6.9999 | 0.0001 | 1158967.1049 | 11.992375001552128 | 23.9995 | 0.0011 | | 963 | 333727.8589 | 3.4994427668141643 | 6.9997 | 0.0003 | 1139917.6212 | 11.953082033429096 | 23.9998 | 0.0008 | | 181 | 334323.6059 | 3.4942578847801946 | 6.9999 | 0.0003 | 1145948.0178 | 11.977131815046297 | 23.9980 | 0.0002 | | 1383 | 334298.1724 | 3.4923130291254028 | 6.9999 | 0.0002 | 1145319.9025 | 11.964814492708198 | 23.9992 | 0.0015 | | 1116 | 337661.6147 | 3.4923165958195015 | 6.9999 | 0.0003 | 1161142.4618 | 12.009292477789144 | 23.9996 | 0.0003 | | 488 | 336083.8006 | 3.5025147266948111 | 7.0000 | 0.0008 | 1149952.2337 | 11.984286735448883 | 23.9992 | 0.0004 | | 121 | 336465.6749 | 3.486221285215461 | 6.9999 | 0.0001 | 1154632.2251 | 11.963489116492081 | 23.9998 | 0.0001 | | 1338 | 337204.2797 | 3.478592072170585 | 7.0000 | 0.0001 | 1162678.8747 | 11.994170179601195 | 23.9996 | 0.0015 | | 910 | 338799.1540 | 3.4927387758889044 | 6.9999 | 0.0001 | 1166834.3104 | 12.029095683549643 | 23.9996 | 0.0001 | | 438 | 335488.9579 | 3.5050823580421069 | 6.9999 | 0.0002 | 1150602.4234 | 12.021129639032544 | 23.9982 | 0.0005 | | 1560 | 338661.5694 | 3.4991844580143199 | 6.9997 | 0.0001 | 1154747.3055 | 11.93130307492018 | 23.9994 | 0.0019 | +-----------+--------------+--------------------------+--------------+--------------+--------------+--------------------------+--------------+--------------+ 1560 tuples (3m 55s) !4 columns dropped! note: to disable dropping columns and/or truncating fields use \w-1 Hear, Monetdb gives only 9 column in result table and gives message 4 columns dropped as i show in following query it gives result of only yellow highlighted statement / columns result. Monetdb can not show blue highlighted part's result. **sql>select product_id, sum(unit_sales),avg(unit_sales), max(unit_sales), min(unit_sales),sum(store_sales),avg(store_sales), max(store_sales), min(store_sales), sum(store_cost),avg(store_cost), max(store_cost), min(store_cost) from sales_fact_1997 group by product_id; my question is that, Monetdb calculate result of all SQL statement or it does not calculate dropped column value. My Next Question How will I see this Dropped result? I hope you will give me proper guidance. My Monetdb version : MonetDB-11.11.07 * * * -- Regards, Swapnil K. Joshi _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

3 2

Problems installing monet db in mountain lion
by Ivan Videla 03 Dec '12

03 Dec '12

Hello, i try to install monetdb under mac os x mountain lion. I downloaded monetdb for mac os x. X86_64 version. I copied monetdb folder to /usr/local, but i haven't been able to start a database. I have followed the instructions from here http://www.monetdb.org/Documentation/monetdbd when i get to this point ( monetdb release my-first-db ) after i try to do a monetdb status. I get that my-first-db is stopped. After i checked log file i get a message that the database appears to shutdown itself after start. I would like to know if someone has running succesfully monetdb in mac os x mountain lion. Regards, Ivan _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

2 2

Can each time it is accessing BAT from disk or from main memory? In my own UDF (xmin)
by swapnil joshi 02 Dec '12

02 Dec '12

Hi I have done some fun with Monetdb. i have created my own user define xmin function. objective of that function is find minimum value from specific column. I have writen some code as follow : /****************************************************************************************/ /* Function Name : xmin */ /****************************************************************************************/ #define UI sht #define UU unsigned short #define UO sht #include "impl_xmin.h" #undef UI #undef UU #undef UO #define UI int #define UU unsigned short #define UO int #include "impl_xmin.h" #undef UI #undef UU #undef UO #define UI lng #define UU unsigned short #define UO lng #include "impl_xmin.h" #undef UI #undef UU #undef UO #define UI dbl #define UU unsigned short #define UO dbl #include "impl_xmin.h" #undef UI #undef UU #undef UO static str UDFBATxmin_(BAT **ret, BAT *bone) { BAT *bres = NULL; bit two_tail_sorted_unsigned = FALSE; bit two_tail_revsorted_unsigned = FALSE; BUN n; str msg = NULL; /* assert calling sanity */ assert(ret != NULL); /* handle NULL pointer */ if (bone == NULL) throw(MAL, "batudf.xmin", RUNTIME_OBJECT_MISSING); n = BATcount(bone); /* allocate result BAT */ switch (bone->ttype) { case TYPE_sht: bres = BATnew(TYPE_void, TYPE_sht, TOTALTHREAD); break; case TYPE_int: bres = BATnew(TYPE_void, TYPE_int, TOTALTHREAD); break; case TYPE_lng: bres = BATnew(TYPE_void, TYPE_lng, TOTALTHREAD); break; case TYPE_dbl: bres = BATnew(TYPE_void, TYPE_dbl, TOTALTHREAD); break; default: throw(MAL, "batudf.xmin","tails of input BATs must be one of {sht, int, lng, dbl}"); } if (bres == NULL) throw(MAL, "batudf.xmin", MAL_MALLOC_FAIL); /* advice on sequential scan */ BATaccessBegin(bone, USE_TAIL, MMAP_SEQUENTIAL); /* call type-specific core algorithm */ switch (bone->ttype) { case TYPE_sht: msg = UDFBATxmin_sht_sht ( bres, bone, n,&two_tail_sorted_unsigned, &two_tail_revsorted_unsigned ); break; case TYPE_int: msg = UDFBATxmin_int_int ( bres, bone, n,&two_tail_sorted_unsigned, &two_tail_revsorted_unsigned ); break; case TYPE_lng: msg = UDFBATxmin_lng_lng ( bres, bone, n,&two_tail_sorted_unsigned, &two_tail_revsorted_unsigned ); break; case TYPE_dbl: msg = UDFBATxmin_dbl_dbl ( bres, bone, n,&two_tail_sorted_unsigned, &two_tail_revsorted_unsigned ); break; default: BBPreleaseref(bres->batCacheid); throw(MAL, "batudf.xmin","tails of input BATs must be one of {sht, int, lng, dbl}"); } BATaccessEnd(bone, USE_TAIL, MMAP_SEQUENTIAL); if (msg != MAL_SUCCEED) { BBPreleaseref(bres->batCacheid); } else { /* set number of tuples in result BAT */ BATsetcount(bres, 1); /* set result properties */ bres->hdense = TRUE; /* result head is dense */ BATseqbase(bres, bone->hseqbase); /* result head has same seqbase as input */ bres->hsorted = 1; /* result head is sorted */ bres->hrevsorted = (BATcount(bres) <= 1); BATkey(bres, TRUE); /* result head is key (unique) */ /* Result tail is sorted, if the left/first input tail is * sorted and key (unique), or if the left/first input tail is * sorted and the second/right input tail is sorted and the * second/right tail values are either all >= 0 or all < 0; * otherwise, we cannot tell. */ if (BATtordered(bone) && (BATtkey(bone) || two_tail_sorted_unsigned)) bres->tsorted = 1; else bres->tsorted = (BATcount(bres) <= 1); if (BATtrevordered(bone) && (BATtkey(bone) || two_tail_revsorted_unsigned)) bres->trevsorted = 1; else bres->trevsorted = (BATcount(bres) <= 1); /* result tail is key (unique), iff both input tails are */ /*BATkey(BATmirror(bres), BATtkey(bone) || BATtkey(btwo));*/ *ret = bres; } return msg; } /* MAL wrapper */ str UDFBATxmin(bat *ires, bat *ione) { // startTimer(); BAT *bres = NULL, *bone = NULL; str msg = NULL; /* assert calling sanity */ assert(ires != NULL && ione != NULL); /* bat-id -> BAT-descriptor */ if ((bone = BATdescriptor(*ione)) == NULL) throw(MAL, "batudf.xmin", RUNTIME_OBJECT_MISSING); /* do the work */ msg = UDFBATxmin_ ( &bres, bone); /* release input BAT-descriptors */ BBPreleaseref(bone->batCacheid); if (msg == MAL_SUCCEED) { /* register result BAT in buffer pool */ BBPkeepref((*ires = bres->batCacheid)); } // printf("Total time Required to UDFBATmin = %lf",endTimer()); return msg; } /****************************************************************************************/ /* Function Name : impl_xmin.h */ /****************************************************************************************/ * impl_xmin.h * * Created on: 23-Nov-2012 * Author: swapnil */ /* concatenate two or five tokens */ #define U_CONCAT_2(a,b) a##b #define U_CONCAT_5(a,b,c,d,e) a##b##c##d##e /* function names, *_nil & TYPE_* macros */ #define UF(p,i,o,s) U_CONCAT_5(p,i,_,o,s) #define UN(t) U_CONCAT_2(t,_nil) #define UT(t) U_CONCAT_2(TYPE_,t) static str UF(UDFBATxmin_,UI,UO,) ( BAT *bres, BAT *bone, BUN n, bit *two_tail_sorted_unsigned, bit *two_tail_revsorted_unsigned ) { UI *one = NULL; UO *res = NULL; BUN i; /* assert calling sanity */ assert(bres != NULL && bone != NULL); assert(BATcapacity(bres) >= 1); assert(BATcount(bone) >= n); assert(bone->ttype == UT(UI)); assert(bres->ttype == UT(UO)); /* get direct access to the tail arrays */ one = (UI*) Tloc(bone, BUNfirst(bone)); res = (UO*) Tloc(bres, BUNfirst(bres)); int loop; startTimer(); res[0]=one[0]; for(loop=1;loop<n;loop++) { if(res[0]>one[loop]) res[0]=one[loop]; } printf("Time for clReleaseMemObject = %lf\n",endTimer()); return MAL_SUCCEED; } /* undo local defines */ #undef UT #undef UN #undef UF #undef U_CONCAT_5 #undef U_CONCAT_2 //*************************************************************************************************/ when i am executing following SQL statement then it will give result but it takes 10 to 13 seconds I know where i am executing on single core of CPU but when i executing that above statement then, I think it will taking data from disk because disk access indicator led blow when i executing that above SQL statement. means, It was not using system main memory. I will be wrong. Can each time it is accessing BAT from disk or from main memory? Can you test me Program ? and Can you give me guidance for accessing BAT? I hope you will give me proper guidance. My SQL statement results are as follow sql>select xmin(v1) from demo; +---------+ | xmin_v1 | +=========+ | 1 | +---------+ 1 tuple (3.2s) sql>select xmin(v1) from demo; +---------+ | xmin_v1 | +=========+ | 1 | +---------+ 1 tuple (9.3s) sql>select xmin(v1) from demo; +---------+ | xmin_v1 | +=========+ | 1 | +---------+ 1 tuple (14.3s) sql>select xmin(v1) from demo; +---------+ | xmin_v1 | +=========+ | 1 | +---------+ 1 tuple (3.4s) -- Regards, Swapnil K. Joshi _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

3 2

Re: users-list Digest, Vol 4, Issue 1
by swapnil joshi 02 Dec '12

02 Dec '12

Hi Lefteris, You have asked me, which version of monetdb are you using? I am using* MonetDB-11.11.7 *version of monetdb. and my dbfarm is on my own system where i have executed this query. My System configuration is as follow: Hardware : Intel i7 2.3GHz Processor, DDR3 8GB RAM, 1TB hard disk S/W : Ubuntu 12.04, MonetDB-11.11.7 On Sat, Dec 1, 2012 at 4:30 PM, <users-list-request(a)monetdb.org> wrote: > Send users-list mailing list submissions to > users-list(a)monetdb.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://mail.monetdb.org/mailman/listinfo/users-list > or, via email, send a message with subject or body 'help' to > users-list-request(a)monetdb.org > > You can reach the person managing the list at > users-list-owner(a)monetdb.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of users-list digest..." > > > Today's Topics: > > 1. Same SQL Statement with different throughput time. (swapnil joshi) > 2. Re: Same SQL Statement with different throughput time. > (Charalampos Nikolaou) > 3. Re: Same SQL Statement with different throughput time. (Lefteris) > 4. Re: Same SQL Statement with different throughput time. > (Stefan Manegold) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 30 Nov 2012 16:59:57 +0530 > From: swapnil joshi <webmaster.swapnil(a)gmail.com> > To: Stefan.Manegold(a)cwi.nl, users-list(a)monetdb.org > Subject: Same SQL Statement with different throughput time. > Message-ID: > < > CAFBccRCGURekT6sCL3SSvOGMWCBwaJv+kFD_bNnyVWEkmC8f3Q(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Stefan, > If monetdb doesn't maintain any caching then when i am creating test1 > database > which contain demo table which contain 100 million records. > > When first time i am executing "SELECT SUM(V1) FROM DEMO;" > monetdb gives me result in (11.4s) > when immediate i am trying to execute same query "SELECT SUM(V1) FROM > DEMO;" > It gives me result in (31.065ms) > > How / why this happen? > > I have also checked my OS memory cache and I have erased OS Memory cache. > Now, I am trying same query "SELECT SUM(V1) FROM DEMO;" > now also he take same time in milliseconds approximate(30.723ms) > > why he first time take lot of time and every next time he take negligible > time. > > I hope you will clear my problem related of throughput time. > > -- > Regards, > Swapnil K. Joshi > _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

1 0

Octopus and
by Masood Mortazavi 02 Dec '12

02 Dec '12

Hi, Where would I find any documentation or instructions to try MonetDB's Octopus functionality, assuming it is still active at (or near) the tip? Regards, Masood P.S. I would also be interested in exploring Monet's any other distributed processing functionality, say Monet/Datacyclotron or Monet/MapReduce. _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

1 1

Re: Same SQL Statement with different throughput time.
by Stefan Manegold 30 Nov '12

30 Nov '12

I guess it's a good idea to run your query in both cases, slow & fast prefixed with TRACE to see where time is spent. BTW, are we talking about genuine MonetDB (as Letters asked: which version) and hence original MonetDB SUM()-implementation, or has the code been modified or extended; f the latter: where & how? S. -- | Stefan.Manegold @ CWI.nl | DB Architectures (INS1) | | http://CWI.nl/~manegold/ | Science Park 123 (L321) | | Tel.: +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) | ----- Reply message ----- From: "Lefteris" <lsidir(a)gmail.com> To: "Communication channel for MonetDB users" <users-list(a)monetdb.org> Subject: Same SQL Statement with different throughput time. Date: Fri, Nov 30, 2012 12:58 just restarting monetdb wont do it (OS will still keep in memory the data), you will also need to clean the memory, for example alloc all memory space, write 1's and then free again the memory. Swapnil, which version of monetdb are you using? IMO, 11 seconds vs. 30 milliseconds is too much of a difference to be just I/O. But maybe I am wrong, maybe your dbfarm is on a network? On Fri, Nov 30, 2012 at 12:42 PM, Charalampos Nikolaou <charnik(a)di.uoa.gr>wrote: > I think the reason is that for the first query the time is spent > mainly for loading your table in main memory. > Then, the second query has only the overhead of computing the sum over > column V1. > > To see this, you can run the query, restart monetdbd, and then run the > query again. You should get similar performance to that of the first > run. > > Babis > > On Fri, Nov 30, 2012 at 1:29 PM, swapnil joshi > <webmaster.swapnil(a)gmail.com> wrote: > > Hi Stefan, > > If monetdb doesn't maintain any caching then when i am creating > test1 > > database > > which contain demo table which contain 100 million records. > > > > When first time i am executing "SELECT SUM(V1) FROM DEMO;" > > monetdb gives me result in (11.4s) > > when immediate i am trying to execute same query "SELECT SUM(V1) FROM > DEMO;" > > It gives me result in (31.065ms) > > > > How / why this happen? > > > > I have also checked my OS memory cache and I have erased OS Memory cache. > > Now, I am trying same query "SELECT SUM(V1) FROM DEMO;" > > now also he take same time in milliseconds approximate(30.723ms) > > > > why he first time take lot of time and every next time he take negligible > > time. > > > > I hope you will clear my problem related of throughput time. > > > > -- > > Regards, > > Swapnil K. Joshi > > > > > > _______________________________________________ > > users-list mailing list > > users-list(a)monetdb.org > > http://mail.monetdb.org/mailman/listinfo/users-list > > > _______________________________________________ > users-list mailing list > users-list(a)monetdb.org > http://mail.monetdb.org/mailman/listinfo/users-list > _______________________________________________ users-list mailing list users-list(a)monetdb.org http://mail.monetdb.org/mailman/listinfo/users-list

1 0