[SERVER-15262] The flush thread should be able to operate with S-lock except for remap Created: 15/Sep/14  Updated: 02/Aug/18  Resolved: 29/Oct/14

Status: Closed
Project: Core Server
Component/s: Storage
Affects Version/s: None
Fix Version/s: 2.8.0-rc0

Type: Task Priority: Major - P3
Reporter: Kaloian Manassiev Assignee: Kaloian Manassiev
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Tested
Participants:

 Description   

The flush thread currently does all its work under an exclusive flush lock. Once deadlock is implemented, this should be changed to be done under S-lock, so that concurrent readers are allowed.



 Comments   
Comment by Githook User [ 06/Nov/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 On long lock acquisitions, dump the call stack as well

Also checks for local deadlock without failing and reduces the output to
only locks which have requests on them.
Branch: master
https://github.com/mongodb/mongo/commit/88ef4861b03af893e59e7339297eea4b9a14bea5

Comment by Githook User [ 05/Nov/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Dump lock manager contents after 30 sec wait, not 5 sec
Branch: master
https://github.com/mongodb/mongo/commit/068eca37eb5b42671a651d24309b052edb1e0473

Comment by Githook User [ 05/Nov/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Indentation fixes in the LockManager
Branch: master
https://github.com/mongodb/mongo/commit/fc3154713b4e7127a0fa9a32930f29423576b037

Comment by Githook User [ 29/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262/SERVER-15711 Flush thread should have precedence and discover deadlocks
Branch: master
https://github.com/mongodb/mongo/commit/e192dc12571f5388a5005c3e71f6c1f497efa6ad

Comment by Githook User [ 29/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Deadlock detection support

Implements the wait cycle detection logic, but it's not incorporated with
the LockerImpl yet.

Also fixes conversion requests to follow same FIFO pattern as conflicting.
Branch: master
https://github.com/mongodb/mongo/commit/ed0947b4ca2c71f4c16fa38cc18a858249910b87

Comment by Githook User [ 29/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Map/reduce incremental collection code should be called without lock
Branch: master
https://github.com/mongodb/mongo/commit/9ee6312856015289fae7c85999f4bbd13fae4fb2

Comment by Githook User [ 28/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Very long lock acquisitions should put warning in the log
Branch: master
https://github.com/mongodb/mongo/commit/083474cc37be86d9cb6302b8fe899293eb692429

Comment by Githook User [ 28/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Chunk migration should yield the MMAP V1 flush lock
Branch: master
https://github.com/mongodb/mongo/commit/bd4479561c02e049ccac63f753b1d28f406583df

Comment by Githook User [ 28/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Remove syncDataAndTruncateJournal from the public interface
Branch: master
https://github.com/mongodb/mongo/commit/b9ebf15a7afa4f8db780b6ba06fc796c1ee2816e

Comment by Githook User [ 28/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Do not reuse write units of work
Branch: master
https://github.com/mongodb/mongo/commit/ca1338ccef5375923da9938e84dc0f7fe393af2c

Comment by Githook User [ 21/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Yield the flush lock even if it is for the LocalDB
Branch: master
https://github.com/mongodb/mongo/commit/57f036a36923b3d6bef6219768f6c65f18bd133a

Comment by Githook User [ 21/Oct/14 ]

Author:

{u'username': u'erh', u'name': u'Eliot Horowitz', u'email': u'eliot@10gen.com'}

Message: SERVER-15262: fsynclock thread after downgrade to global S should unlock flush lock as it doesn't need it
Branch: master
https://github.com/mongodb/mongo/commit/25a2dee43f03140c197df7e9e5e4a4e8e57a97cb

Comment by Githook User [ 08/Oct/14 ]

Author:

{u'username': u'kaloianm', u'name': u'Kaloian Manassiev', u'email': u'kaloian.manassiev@mongodb.com'}

Message: SERVER-15262 Lock manager fixes and unit-tests

o Timeout of a lock upgrade throws an assertion error
o Add better tracing to the LockManager dump method
o Move LockHead inside lock_mgr_new.cpp, since it is not needed outside
Branch: master
https://github.com/mongodb/mongo/commit/0a7a02fda732db87515cd90a7c8689a1344fe311

Generated at Thu Feb 08 03:37:29 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.