MonetDB not cleaning intermediate results which leads to filling up disk space
Hi, Right now I am evaluating MonetDB and try to analyze a dataset that consists of 10 million rows. The data set has 2 tables, table a stores 10 million rows, table b stores 1 million rows. Both tables are being joined and some statistical values are being calculated (QUARTILE, AVG, SUM and GROUP BY). MonetDB is promising fast here. When starting up the database and executing the query the result is being returned within 4 seconds. I tried to execute the same SQL query 1000 times sequentially and to my surprise after the 60th iteration monetDB crashed because of insuffcient disk space. As far as I read intermediate results are being materialized within the bat folder. The .tail files are not being removed after a query was executed successfully and the files are not even removed after the database conenction is being closed. When I restart the server (dbfarm) the files are cleaned up perfectly. I am just wondering if this default behaviour can be changed? Are there any ways to enforce cleaning up the files? How to avoid a server crash because of insufficient memory? For the system I am planning to use monetDB for I am not able to restart the server before I execute the SQL queries. Thank you very much, Alex
hi, are you on windows? mserver version?
On Tue, Oct 13, 2015 at 5:46 PM,
Hi,
Right now I am evaluating MonetDB and try to analyze a dataset that consists of 10 million rows. The data set has 2 tables, table a stores 10 million rows, table b stores 1 million rows.
Both tables are being joined and some statistical values are being calculated (QUARTILE, AVG, SUM and GROUP BY).
MonetDB is promising fast here. When starting up the database and executing the query the result is being returned within 4 seconds.
I tried to execute the same SQL query 1000 times sequentially and to my surprise after the 60th iteration monetDB crashed because of insuffcient disk space.
As far as I read intermediate results are being materialized within the bat folder. The .tail files are not being removed after a query was executed successfully and the files are not even removed after the database conenction is being closed.
When I restart the server (dbfarm) the files are cleaned up perfectly. I am just wondering if this default behaviour can be changed? Are there any ways to enforce cleaning up the files? How to avoid a server crash because of insufficient memory?
For the system I am planning to use monetDB for I am not able to restart the server before I execute the SQL queries.
Thank you very much,
Alex
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
Hi,
I am using Ubuntu 14.04 64 bit (but I have seen the same result for MacOS) and the server version is
# MonetDB 5 server v11.21.5 "Jul2015"
# Serving database 'demo', using 4 threads
# Compiled for x86_64-pc-linux-gnu/64bit with 64bit OIDs and 128bit integers dynamically linked
# Found 3.847 GiB available main-memory.
# Copyright (c) 1993-July 2008 CWI.
# Copyright (c) August 2008-2015 MonetDB B.V., all rights reserved
Thanks,
Alex
Anthony Damico
This is happening for us as well on Ubuntu 14.04 (actually in a ubuntu-based Docker image). See attached screenshot showing the disk usage on our MonetDB server. The drop in disk usage occurs when the dbfarm is restarted nightly (for a backup).
Hi Dennis,
thank you very much for sharing your experience as well.
As you face the same issue and as we are not able to restart the server several times per day I believe that we might have to look for a different dbms.
I would consider this as a serious bug and I am just wondering if there is a way to open a defect and if someone is going to fix this in a future release.
Thanks again,
Alex
Dennis Pallett
Dear Alex et al., Sorry for the inconveniences. Please file a bug report via http://bugs.monetdb.org . For commercial support, you might also want to consider contacting MonetDB Solutions (cf., http://www.monetdbsolutions.com). Best, Stefan On October 15, 2015 6:07:24 AM CEST, abraun_75@yahoo.de wrote:
Hi Dennis, thank you very much for sharing your experience as well. As you face the same issue and as we are not able to restart the server several times per day I believe that we might have to look for a different dbms. I would consider this as a serious bug and I am just wondering if there is a way to open a defect and if someone is going to fix this in a future release. Thanks again, Alex
Dennis Pallett
schrieb am 4:03 Mittwoch, 14.Oktober 2015: This is happening for us as well on Ubuntu 14.04 (actually in a ubuntu-based Docker image). See attached screenshot showing the disk usage on our MonetDB server. The drop in disk usage occurs when the dbfarm is restarted nightly (for a backup). _______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
------------------------------------------------------------------------
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
-- | Stefan.Manegold@CWI.nl | DB Architectures (DA) | | www.CWI.nl/~manegold/ | Science Park 123 (L321) | | +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) |
Hello Alex (and Dennis), Are you still have this problem? When was your MonetDB (compiled and) installed (source or using installer)? We’ve recently (after the Jul2015 release) fixed some leak in MonetDB, e.g., this: http://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=773e781c6a1b Would you please refresh your MonetDB installation (you’ll need to install from our Mercurial source) and try if your problem persists? And indeed, a bug report with script/information/data to reproduce this problem is highly appreciated. For mission critical problems, you might want to consider the services of MonetDB Solutions. Regards, Jennie
On Oct 15, 2015, at 07:08, Stefan Manegold
wrote: Dear Alex et al.,
Sorry for the inconveniences. Please file a bug report via http://bugs.monetdb.org .
For commercial support, you might also want to consider contacting MonetDB Solutions (cf., http://www.monetdbsolutions.com).
Best, Stefan
On October 15, 2015 6:07:24 AM CEST, abraun_75@yahoo.de wrote: Hi Dennis,
thank you very much for sharing your experience as well.
As you face the same issue and as we are not able to restart the server several times per day I believe that we might have to look for a different dbms.
I would consider this as a serious bug and I am just wondering if there is a way to open a defect and if someone is going to fix this in a future release.
Thanks again,
Alex
Denn! is Pallett
schrieb am 4:03 Mittwoch, 14.Oktober 2015: This is happening for us as well on Ubuntu 14.04 (actually in a ubuntu-based Docker image). See attached screenshot showing the disk usage on our MonetDB server. The drop in disk usage occurs when the dbfarm is restarted nightly (for a backup).
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
-- | Stefan.Manegold@CWI.nl | DB Architectures (DA) | | www.CWI.nl/~manegold/ | Science Park 123 (L321) | | +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) | _______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
Hi Jennie,
for this issue I have created a defect
https://www.monetdb.org/bugzilla/show_bug.cgi?id=3825
This defect has already been resolved by Niels.
Yesterday I have downloaded the latest sources that included the fix implemented by Niels. I compiled the sources and executed the same queries (2000 times). I can confirm the issue is resolved for me, the .tail files are being deleted after the query has been executed. This is fantastic, thank you very much for the super fast fix!
Will there be a new official release or service pack in the near future?
Thanks,
Alex
Ying Zhang
On Oct 15, 2015, at 07:08, Stefan Manegold
wrote: Dear Alex et al.,
Sorry for the inconveniences. Please file a bug report via http://bugs.monetdb.org .
For commercial support, you might also want to consider contacting MonetDB Solutions (cf., http://www.monetdbsolutions.com).
Best, Stefan
On October 15, 2015 6:07:24 AM CEST, abraun_75@yahoo.de wrote: Hi Dennis,
thank you very much for sharing your experience as well.
As you face the same issue and as we are not able to restart the server several times per day I believe that we might have to look for a different dbms.
I would consider this as a serious bug and I am just wondering if there is a way to open a defect and if someone is going to fix this in a future release.
Thanks again,
Alex
Denn! is Pallett
schrieb am 4:03 Mittwoch, 14.Oktober 2015: This is happening for us as well on Ubuntu 14.04 (actually in a ubuntu-based Docker image). See attached screenshot showing the disk usage on our MonetDB server. The drop in disk usage occurs when the dbfarm is restarted nightly (for a backup).
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
-- | Stefan.Manegold@CWI.nl | DB Architectures (DA) | | www.CWI.nl/~manegold/ | Science Park 123 (L321) | | +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) | _______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
On Oct 19, 2015, at 01:21, abraun_75@yahoo.de wrote:
Hi Jennie,
for this issue I have created a defect
https://www.monetdb.org/bugzilla/show_bug.cgi?id=3825
This defect has already been resolved by Niels.
Yesterday I have downloaded the latest sources that included the fix implemented by Niels. I compiled the sources and executed the same queries (2000 times). I can confirm the issue is resolved for me, the .tail files are being deleted after the query has been executed. This is fantastic, thank you very much for the super fast fix!
Thanks for the information.
Will there be a new official release or service pack in the near future?
Sure. See Sjoerd’s reply yesterday evening to another thread. He plans to build a RC for Jul2015-SP1 this week. Jennie
Thanks,
Alex
Ying Zhang
schrieb am 9:14 Sonntag, 18.Oktober 2015: Hello Alex (and Dennis),
Are you still have this problem? When was your MonetDB (compiled and) installed (source or using installer)? We’ve recently (after the Jul2015 release) fixed some leak in MonetDB, e.g., this: http://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=773e781c6a1b
Would you please refresh your MonetDB installation (you’ll need to install from our Mercurial source) and try if your problem persists?
And indeed, a bug report with script/information/data to reproduce this problem is highly appreciated.
For mission critical problems, you might want to consider the services of MonetDB Solutions.
Regards,
Jennie
On Oct 15, 2015, at 07:08, Stefan Manegold
wrote: Dear Alex et al.,
Sorry for the inconveniences. Please file a bug report via http://bugs.monetdb.org .
For commercial support, you might also want to consider contacting MonetDB Solutions (cf., http://www.monetdbsolutions.com).
Best, Stefan
On October 15, 2015 6:07:24 AM CEST, abraun_75@yahoo.de wrote: Hi Dennis,
thank you very much for sharing your experience as well.
As you face the same issue and as we are not able to restart the server several times per day I believe that we might have to look for a different dbms.
I would consider this as a serious bug and I am just wondering if there is a way to open a defect and if someone is going to fix this in a future release.
Thanks again,
Alex
Denn! is Pallett
schrieb am 4:03 Mittwoch, 14.Oktober 2015: This is happening for us as well on Ubuntu 14.04 (actually in a ubuntu-based Docker image). See attached screenshot showing the disk usage on our MonetDB server. The drop in disk usage occurs when the dbfarm is restarted nightly (for a backup).
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
-- | Stefan.Manegold@CWI.nl | DB Architectures (DA) | | www.CWI.nl/~manegold/ | Science Park 123 (L321) | | +31 (0)20 592-4212 | 1098 XG Amsterdam (NL) |
_______________________________________________ users-list mailing list users-list@monetdb.org https://www.monetdb.org/mailman/listinfo/users-list
participants (5)
-
abraun_75@yahoo.de
-
Anthony Damico
-
Dennis Pallett
-
Stefan Manegold
-
Ying Zhang