summaryrefslogtreecommitdiff
path: root/src/rabbit_msg_store.erl
Commit message (Collapse)AuthorAgeFilesLines
* Remove a couple of TODOs, make the read buffer size configurable, and don't ↵bug26457Simon MacMullen2014-11-131-1/+2
| | | | use the read buffer for the QI or msg store transform since they already read in decent sized chunks.
* Switch to Erlang 17-style types, /0 versions at least.Simon MacMullen2014-09-171-2/+2
|
* Update copyright for 2014bug25940Simon MacMullen2014-03-171-1/+1
|
* correctness++bug26066Matthias Radestock2014-03-131-1/+1
|
* turns out that cleaning up 'clients' in client_terminate is badMatthias Radestock2014-03-131-6/+6
| | | | | since it breaks clean shutdown/recovery, which relies on the CRefs in 'clients'.
* invoke credit_flow:peer_down on, err, DOWN. *only*Matthias Radestock2014-03-131-8/+11
| | | | | | | | Thus preventing potential leaks and deadlocks. Also, clean up 'clients' in clear_client, and thus client_terminate, which was previously missing and thus the source of another potential leak.
* s/VMware/GoPivotal/gSimon MacMullen2013-07-011-2/+2
|
* cosmetic: state the not-immediatly-obviousMatthias Radestock2013-06-171-1/+8
|
* Throw the key too. And if the server side still fails try to fail with more ↵bug25611Simon MacMullen2013-06-171-2/+3
| | | | information.
* Check value when clients update flying_etsEmile Joubert2013-06-171-1/+4
|
* Merged stable into defaultEmile Joubert2013-04-101-13/+10
|\
| * Merge in defaultSimon MacMullen2013-01-301-9/+6
| |\
| | * stable to defaultSimon MacMullen2013-01-241-1/+1
| | |\
| | * | rationalise timer maintenanceMatthias Radestock2013-01-081-8/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - introduce a couple of helper functions and use them wherever we can - pay attention to the result of timer cancellation, to prevent duplicate timer creation - stop all timers when a queue terminates. More out of politeness than necessity.
| * | | Pass length to prioritise_*Simon MacMullen2012-11-081-5/+5
| |/ /
* | | Only queue up for GC at most two pairs of files at once, since we lock files ↵bug25514Simon MacMullen2013-04-051-3/+8
| |/ |/| | | | | as soon as they are queued; no point in having them locked for ages as they wait to get GCed.
* | Update copyright 2013bug25343Emile Joubert2013-01-231-1/+1
|/
* file_name -> filenameMatthias Radestock2012-09-171-5/+5
| | | | | | | | | The latter is used in more places and also by OTP. In the process fix a bug - rabbit_upgrade was referencing rabbit_node_monitor:running_nodes_filename/0, which was in fact named running_nodes_file_name. That function was also missing a spec. And it was in the "wrong" place w.r.t. the function order in the module ;)
* cosmeticMatthias Radestock2012-04-111-2/+2
|
* Changed uses of gb_sets:difference/2 to our difference functions.Francesco Mazzoli2012-04-111-1/+2
|
* Update copyright 2012bug24606Emile Joubert2012-02-021-1/+1
|
* More credit for the queue -> msg_store case.Simon MacMullen2012-01-191-2/+2
|
* two flavours of 'write'Matthias Radestock2012-01-171-11/+19
| | | | | | | Just like in the other applications of credit_flow:send/ack, we need two flavours of the surrounding API function, i.e. 'write' and 'write_flow' in this case. That way existing call sites, such as in rabbit_test can remain undisturbed.
* more sensible arg orderMatthias Radestock2012-01-171-13/+13
| | | | and consistent var naming
* Don't pass the client pid in for write.Simon MacMullen2012-01-171-3/+5
|
* Flow control: vq -> msg_store.Simon MacMullen2012-01-171-17/+26
|
* more accurate description of flying_ets tablebug24308Matthias Radestock2011-10-171-1/+1
|
* explain why client_update_flying phase 3 will always succeedMatthias Radestock2011-10-171-0/+3
|
* refactor: revert change to var nameMatthias Radestock2011-10-171-2/+2
|
* refactor: rename update_flying/3 to client_update_flying/3Matthias Radestock2011-10-171-4/+4
|
* fix bugMatthias Radestock2011-10-171-1/+1
|
* Give up on attempting to improve clarity.Simon MacMullen2011-10-171-6/+6
|
* Rephrase to make Matthias happier...Simon MacMullen2011-10-171-7/+7
|
* I find this clearerSimon MacMullen2011-10-171-6/+6
|
* rename 'removed' confirmation event name to 'ignored'Matthias Radestock2011-10-151-3/+3
| | | | since that reflects its meaning better
* merge default into bug24308Matthias Radestock2011-10-151-13/+8
|\
| * a spot of inliningMatthias Radestock2011-10-151-14/+7
| |
| * only confirm messages on 'remove' that are pending confirmationMatthias Radestock2011-10-111-2/+3
| |
* | documentationMatthias Radestock2011-10-141-0/+39
| |
* | issue confirms when needed (and only then)Matthias Radestock2011-10-111-22/+33
| | | | | | | | | | | | This also fixes a bug that is present on 'default', namely that we used to always issue a confirm on 'remove' regardless of whether there was preceding 'write'.
* | refactoring: extract core of 'write' handlerMatthias Radestock2011-10-101-66/+62
| | | | | | | | | | | | similar to what we already did for 'remove' Also, don't add the 'dying client' marker message to the pending confirms. While that does no harm, it is clearly not right.
* | cosmeticMatthias Radestock2011-10-101-2/+2
| |
* | there's more than one call site for remove_message...Matthias Radestock2011-10-101-41/+41
| | | | | | | | ...and only one of them must pay attention to flying_ets
* | be more careful about deleting cur_file_cache_ets entriesMatthias Radestock2011-10-101-6/+17
| |
* | I'm pretty sure this is correctMatthias Radestock2011-10-091-31/+47
| |
* | slightly more obvious flying_writes accountingMatthias Radestock2011-10-061-14/+11
| | | | | | | | | | and move the cache cleanup from 'remove' to 'write', which is a more obvious place. Improve docs on that too.
* | merge default into bug24308Matthias Radestock2011-10-061-73/+116
|\ \ | |/ |/|
| * Whoops, forgot that bit.Matthew Sackman2011-08-191-0/+4
| |
| * Fixed again.Matthew Sackman2011-08-191-73/+112
| |
* | move file functions from misc to rabbit_fileAlexandru Scvortov2011-09-211-4/+4
| |
View Lorry Controller Status