summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Merge bug26543Simon MacMullen2015-01-211-109/+87
|\
| * Fix a couple of silly bugs: specify the journal prefix lengths for ack and ↵Simon MacMullen2015-01-211-18/+22
| | | | | | | | del, and correctly assemble the segment backwards.
| * Also, switch to reading the journal in one go. We don't care about this for ↵Simon MacMullen2015-01-201-44/+27
| | | | | | | | performance, but it's probably a simplification.
| * Read segment files in one call to fhc:read/2.Simon MacMullen2015-01-201-35/+27
| |
| * Write segment files in one append/2 operation.Simon MacMullen2015-01-201-30/+29
|/
* Merge bug26542 (again)Simon MacMullen2015-01-201-1/+4
|\
| * Oops, headers can be undefined.Simon MacMullen2015-01-201-1/+4
| |
* | Merge bug26542Simon MacMullen2015-01-200-0/+0
|\ \ | |/ |/|
* | Improve the performance of determine_persist_to/2, stop it being a hotspot.Simon MacMullen2015-01-202-6/+24
|/
* Merge bug26484.Simon MacMullen2015-01-191-39/+46
|\
| * Allow seeking backwards through the read buffer.Simon MacMullen2015-01-151-39/+46
| |
* | merge bug26426 into defaultMichael Klishin2015-01-169-41/+150
|\ \ | |/ |/|
| * Merge in defaultSimon MacMullen2015-01-0840-282/+1249
| |\
| * | Not needed any moreSimon MacMullen2014-10-271-8/+0
| | |
| * | Start distribution early so we correctly generate node from "-n"Simon MacMullen2014-10-271-4/+2
| | |
| * | Explain what we are doing and why.Simon MacMullen2014-10-271-2/+14
| | |
| * | Check in with epmd periodically, if it hasn't heard of us then re-register ↵Simon MacMullen2014-10-272-0/+96
| | | | | | | | | | | | (and restart it if needed).
| * | Revert to stable-like behaviour, always starting epmd, but starting it ↵Simon MacMullen2014-10-273-11/+1
| | | | | | | | | | | | ourselves since we will need this capability later and it lets us get better error messages.
| * | *Do* start epmd ourselves if we want to impersonate a node. And just allow ↵Simon MacMullen2014-10-243-16/+41
| | | | | | | | | | | | net_kernel:start() to fail and catch the error later.
| * | First pass at having ctl and plugins not start epmd. This can cause problems ↵Simon MacMullen2014-10-236-32/+28
| | | | | | | | | | | | though if we are contacting a remote node, or removing a cluster node offline.
* | | Merge bug26522Simon MacMullen2015-01-135-10/+28
|\ \ \
| * \ \ Merge in defaultSimon MacMullen2015-01-135-10/+28
| |\ \ \ |/ / / /
| * | | Merge bug26327Simon MacMullen2015-01-122-2/+2
| |\ \ \
| * | | | More stats (and rename the existing ones).Simon MacMullen2015-01-075-10/+28
| | | | |
* | | | | Merge bug26327Simon MacMullen2015-01-135-273/+521
|\ \ \ \ \
| * | | | | Fix confusion over where the embedded size lives.Simon MacMullen2015-01-131-8/+7
| | | | | |
| * | | | | Oops.Simon MacMullen2015-01-131-11/+13
| | | | | |
| * | | | | Simplify reading message body.Simon MacMullen2015-01-131-46/+40
| | | | | |
| * | | | | Finish the job started by 2f668f7b457f; make sure we always hold messages in ↵Simon MacMullen2015-01-131-21/+37
| | | | | | | | | | | | | | | | | | | | | | | | memory in the serialised form.
| * | | | | Rename this for a bit more clarity.Simon MacMullen2015-01-132-5/+5
| | | | | |
| * | | | | Only call create_pub_record_body/2 once per message published; keep the ↵Simon MacMullen2015-01-121-10/+8
| | |/ / / | |/| | | | | | | | | | | | | serialised form in memory.
| * | | | So that term_to_binary/1 there somehow seemed to end up creating a lot of ↵Simon MacMullen2015-01-092-2/+2
| |/ / / | | | | | | | | | | | | garbage under heavy load; pumping millions of messages into a queue became almost quadratic. I'm not sure how or why, but let's try another tactic.
| * | | Decide where to persist to based on the encoded size of the message.Simon MacMullen2015-01-061-11/+21
| | | |
| * | | Handle read failure in read_pub_record_bodySimon MacMullen2014-12-121-15/+25
| | | |
| * | | Remove vestigial use of read_ahead.Simon MacMullen2014-12-121-4/+2
| | | |
| * | | Use msg_in_ram/1 more.Simon MacMullen2014-12-121-21/+19
| | | |
| * | | Rephrase, and add a high level introduction.Simon MacMullen2014-12-111-6/+19
| | | |
| * | | Reduce distance to default, add a tiny bit more explanation.Simon MacMullen2014-12-112-16/+10
| | | |
| * | | Merge in defaultSimon MacMullen2014-12-114-248/+464
| |\ \ \
| | * | | Run the VQ tests both while embedding messages in the QI and sending them to ↵Simon MacMullen2014-12-111-5/+9
| | | | | | | | | | | | | | | | | | | | the store.
| | * | | A few fixes to VQ book-keeping, which get the tests passing again:Simon MacMullen2014-12-113-109/+139
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Don't count messages / bytes read during read, do so in betas_from_index_entries since unacked msgs don't count * Introduce qi_pending_ack to prevent us trying to page out messages that don't touch the store * Fix purge bookkeeping, account for the fact messages can be paged in to be purged
| | * | | Clearer test failures.Simon MacMullen2014-12-051-5/+14
| | | | |
| | * | | API changesSimon MacMullen2014-12-051-8/+12
| | | | |
| | * | | Don't recover transient messages.Simon MacMullen2014-12-051-2/+2
| | | | |
| | * | | Update essay.Simon MacMullen2014-12-051-6/+8
| | | | |
| | * | | Introduce a selection mechanism so that small messages go to the index, and ↵Simon MacMullen2014-12-052-13/+34
| | | | | | | | | | | | | | | | | | | | large ones go to the store.
| | * | | Fix bookkeeping of ram_msg_{count,bytes} - don't assume it decrements on ↵Simon MacMullen2014-12-051-23/+28
| | | | | | | | | | | | | | | | | | | | alpha -> beta transition, allow it to decrement on beta -> delta as well if the message went to the QI. (We were already correct on delta -> beta, and there isn't a bulk beta -> alpha move.)
| | * | | Reinstate this.Simon MacMullen2014-12-051-1/+1
| | | | |
| | * | | UpgradeSimon MacMullen2014-12-051-3/+36
| | | | |
| | * | | Don't leak on noack.Simon MacMullen2014-12-041-3/+4
| | | | |
View Lorry Controller Status