New Galaxy won't start, SQLite (OperationalError) database is locked
Hello all, While attempting a fresh Galaxy install from both galaxy-central and galaxy-dist, I ran into a problem initialising the default SQLite database (database/universe.sqlite), $ hg clone https://bitbucket.org/galaxy/galaxy-dist ... $ cd galaxy-dist $ ./run.sh ... galaxy.model.migrate.check DEBUG 2013-10-31 10:28:26,143 pysqlite>=2 egg successfully loaded for sqlite dialect Traceback (most recent call last): ... OperationalError: (OperationalError) database is locked u'PRAGMA table_info("dataset")' () After some puzzlement, I realised this was down to the file system - I was trying this under my home directory mounted via a distributed file system (gluster I think). Repeating the experiment under /tmp on a local hard disk worked :) (I'm posting this message for future reference; hopefully Google and/or mailing list searches will help anyone else facing this error) Regards, Peter
Hi Peter, Thanks for reporting this. I am now facing the same error, but I don't get it.... For example: normally it appears when I try to change something in the DB (e.g. adding a new history), but then suddenly I managed to create one new history item....and now it is occurring again. It seems to be a transient locked state, so every now and then the lock is released somehow. Anyway, are there any ideas on what could be a real solution to this problem? This is the second Galaxy server I upgrade which now is stuck after the upgrade...any help is appreciated! Thanks and regards, Pieter. -----Original Message----- From: galaxy-dev-bounces@lists.bx.psu.edu [mailto:galaxy-dev-bounces@lists.bx.psu.edu] On Behalf Of Peter Cock Sent: donderdag 31 oktober 2013 11:54 To: Galaxy Dev Subject: [galaxy-dev] New Galaxy won't start, SQLite (OperationalError) database is locked Hello all, While attempting a fresh Galaxy install from both galaxy-central and galaxy-dist, I ran into a problem initialising the default SQLite database (database/universe.sqlite), $ hg clone https://bitbucket.org/galaxy/galaxy-dist ... $ cd galaxy-dist $ ./run.sh ... galaxy.model.migrate.check DEBUG 2013-10-31 10:28:26,143 pysqlite>=2 egg successfully loaded for sqlite dialect Traceback (most recent call last): ... OperationalError: (OperationalError) database is locked u'PRAGMA table_info("dataset")' () After some puzzlement, I realised this was down to the file system - I was trying this under my home directory mounted via a distributed file system (gluster I think). Repeating the experiment under /tmp on a local hard disk worked :) (I'm posting this message for future reference; hopefully Google and/or mailing list searches will help anyone else facing this error) Regards, Peter ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
Hi Pieter, Using SQLite is NOT recommended for a production Galaxy, it is only really used for development and testing (having the whole database as one file on disk without worrying about creating a proper database, user accounts etc is nice). Peter On Wednesday, December 18, 2013, Lukasse, Pieter wrote:
Hi Peter,
Thanks for reporting this. I am now facing the same error, but I don't get it.... For example: normally it appears when I try to change something in the DB (e.g. adding a new history), but then suddenly I managed to create one new history item....and now it is occurring again. It seems to be a transient locked state, so every now and then the lock is released somehow.
Anyway, are there any ideas on what could be a real solution to this problem? This is the second Galaxy server I upgrade which now is stuck after the upgrade...any help is appreciated!
Thanks and regards,
Pieter.
-----Original Message----- From: galaxy-dev-bounces@lists.bx.psu.edu <javascript:;> [mailto: galaxy-dev-bounces@lists.bx.psu.edu <javascript:;>] On Behalf Of Peter Cock Sent: donderdag 31 oktober 2013 11:54 To: Galaxy Dev Subject: [galaxy-dev] New Galaxy won't start, SQLite (OperationalError) database is locked
Hello all,
While attempting a fresh Galaxy install from both galaxy-central and galaxy-dist, I ran into a problem initialising the default SQLite database (database/universe.sqlite),
$ hg clone https://bitbucket.org/galaxy/galaxy-dist ... $ cd galaxy-dist $ ./run.sh ... galaxy.model.migrate.check DEBUG 2013-10-31 10:28:26,143 pysqlite>=2 egg successfully loaded for sqlite dialect Traceback (most recent call last): ... OperationalError: (OperationalError) database is locked u'PRAGMA table_info("dataset")' ()
After some puzzlement, I realised this was down to the file system - I was trying this under my home directory mounted via a distributed file system (gluster I think).
Repeating the experiment under /tmp on a local hard disk worked :)
(I'm posting this message for future reference; hopefully Google and/or mailing list searches will help anyone else facing this error)
Regards,
Peter ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
Hi Peter, I know that ;) Still I am curious about the possible solution. Any ideas? Regards, Pieter. From: Peter Cock [mailto:p.j.a.cock@googlemail.com] Sent: woensdag 18 december 2013 13:31 To: Lukasse, Pieter Cc: Galaxy Dev Subject: Re: New Galaxy won't start, SQLite (OperationalError) database is locked Hi Pieter, Using SQLite is NOT recommended for a production Galaxy, it is only really used for development and testing (having the whole database as one file on disk without worrying about creating a proper database, user accounts etc is nice). Peter On Wednesday, December 18, 2013, Lukasse, Pieter wrote: Hi Peter, Thanks for reporting this. I am now facing the same error, but I don't get it.... For example: normally it appears when I try to change something in the DB (e.g. adding a new history), but then suddenly I managed to create one new history item....and now it is occurring again. It seems to be a transient locked state, so every now and then the lock is released somehow. Anyway, are there any ideas on what could be a real solution to this problem? This is the second Galaxy server I upgrade which now is stuck after the upgrade...any help is appreciated! Thanks and regards, Pieter. -----Original Message----- From: galaxy-dev-bounces@lists.bx.psu.edu<javascript:;> [mailto:galaxy-dev-bounces@lists.bx.psu.edu<javascript:;>] On Behalf Of Peter Cock Sent: donderdag 31 oktober 2013 11:54 To: Galaxy Dev Subject: [galaxy-dev] New Galaxy won't start, SQLite (OperationalError) database is locked Hello all, While attempting a fresh Galaxy install from both galaxy-central and galaxy-dist, I ran into a problem initialising the default SQLite database (database/universe.sqlite), $ hg clone https://bitbucket.org/galaxy/galaxy-dist ... $ cd galaxy-dist $ ./run.sh ... galaxy.model.migrate.check DEBUG 2013-10-31 10:28:26,143 pysqlite>=2 egg successfully loaded for sqlite dialect Traceback (most recent call last): ... OperationalError: (OperationalError) database is locked u'PRAGMA table_info("dataset")' () After some puzzlement, I realised this was down to the file system - I was trying this under my home directory mounted via a distributed file system (gluster I think). Repeating the experiment under /tmp on a local hard disk worked :) (I'm posting this message for future reference; hopefully Google and/or mailing list searches will help anyone else facing this error) Regards, Peter ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/ To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
On Wed, Dec 18, 2013 at 12:51 PM, Lukasse, Pieter <pieter.lukasse@wur.nl> wrote:
Hi Peter,
I know that ;) Still I am curious about the possible solution. Any ideas?
Regards,
Pieter.
Could you clarify the nature of the problem - is it failing to even start as in the situation I reported at the start of this thread, or an intermittent problem? Note that while generally stable I have had problems with my development machine with heavy SQLite DB access, e.g. a large multi-part job being split creating many child jobs in quick succession. Peter
Hi Peter, Galaxy starts, but any operation that involves changes in the database seems to fail. One "solution" I found was to move to the default branch (hg update default) instead of the "stable" one...there the problem seems to be fixed somehow. However, then I get another error when trying to create a new history because the attribute "genome_build" in the payload is "null" and, according to the error, it should be a string. Details: source annotation deleted false genome_build null id cc4287507a7e4aae model_class History name test1235 nice_size 0 bytes published false purged false state new state_details discarded 0 empty 0 error 0 failed_metadata 0 new 0 ok 0 paused 0 queued 0 running 0 setting_metadata 0 upload 0 state_ids discarded empty error failed_metadata new ok paused queued running setting_metadata upload tags url /api/histories/cc4287507a7e4aae user username pieterlukasse quota_percent null total_disk_usage 36654268543 nice_total_disk_usage 34.1 GB email pieter.lukasse@wur.nl<mailto:pieter.lukasse@wur.nl> model_class User id f2db41e1fa331b3e diskSize 0 xhr readyState 4 responseText { "error": "genome_build must be a string: " } responseJSON error genome_build must be a string: status 400 statusText Bad Request responseHeaders Date Wed, 18 Dec 2013 13:43:26 GMT cache-control max-age=0,no-cache,no-store Server PasteWSGIServer/0.5 Python/2.7.5+ Connection close content-type application/json options validate true parse true emulateHTTP false emulateJSON false Disabling this specific validation in /home/lukas007/galaxy-dist/lib/galaxy/webapps/galaxy/api/histories.py (by disabling I mean just putting a dummy string value in "genome_build") will restore the "normal" Galaxy functionality and all seems to be working fine now...until I hit something that depends on this "genome_build" parameter of course... Is this a known issue? If not, hereby ;) Regards, Pieter. -----Original Message----- From: Peter Cock [mailto:p.j.a.cock@googlemail.com] Sent: woensdag 18 december 2013 17:36 To: Lukasse, Pieter Cc: Galaxy Dev Subject: Re: New Galaxy won't start, SQLite (OperationalError) database is locked On Wed, Dec 18, 2013 at 12:51 PM, Lukasse, Pieter <pieter.lukasse@wur.nl<mailto:pieter.lukasse@wur.nl>> wrote:
Hi Peter,
I know that ;) Still I am curious about the possible solution. Any ideas?
Regards,
Pieter.
Could you clarify the nature of the problem - is it failing to even start as in the situation I reported at the start of this thread, or an intermittent problem? Note that while generally stable I have had problems with my development machine with heavy SQLite DB access, e.g. a large multi-part job being split creating many child jobs in quick succession. Peter
participants (2)
-
Lukasse, Pieter -
Peter Cock