License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
dont GC blocks used by pinner
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
comment GC algo
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
add lock to blockstore to prevent GC from eating wanted blocks
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
improve FetchGraph
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
separate interfaces for blockstore and GCBlockstore
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
reintroduce indirect pinning, add enumerateChildren dag method
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
* ID service stream
* make the relay service use msmux
* fix nc tests
Note from jbenet: Maybe we should remove the old protocol/muxer
and see what breaks. It shouldn't be used by anything now.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
Signed-off-by: Juan Batiz-Benet <juan@benet.ai>
WARNING: No migration performed! That needs to come in a separate
commit, perhaps amended into this one.
This is the minimal rewrite, only changing the storage from
JSON(+extra keys) in Datastore to IPFS objects. All of the pinning
state is still loaded in memory, and written from scratch on Flush. To
do more would require API changes, e.g. adding error returns.
Set/Multiset is not cleanly separated into a library, yet, as it's API
is expected to change radically.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
Pinner had method GetManual that returned a ManualPinner, so every
Pinner had to implement ManualPinner anyway.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
These secondary copies were never actually queried, and didn't contain
the indirect refcounts so they couldn't become the authoritative
source anyway as is. New goal is to move pinning into IPFS objects.
A migration will be needed to remove the old data from the datastore.
This can happen at any time after this commit.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
This used to lead to large refcount numbers, causing Flush to create a
lot of IPFS objects, and merkledag to consume tens of gigabytes of
RAM.
License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>
Instead put it inside of DAG.Get.
The fix is applied only in the case when the context.WithCancel
before a DAG.Get is also used later on in the scope.
License: MIT
Signed-off-by: rht <rhtbot@gmail.com>
This commit adds a new set of sharness tests for pinning, and addresses
bugs that were pointed out by said tests.
test/sharness: added more pinning tests
Pinning is currently broken. See issue #1051. This commit introduces
a few more pinning tests. These are by no means exhaustive, but
definitely surface the present problems going on. I believe these
tests are correct, but not sure. Pushing them as failing so that
pinning is fixed in this PR.
make pinning and merkledag.Get take contexts
improve 'add' commands usage of pinning
FIXUP: fix 'pin lists look good'
ipfs-pin-stat simple script to help check pinning
This is a simple shell script to help check pinning.
We ought to strive towards making adding commands this easy.
The http api is great and powerful, but our setup right now
gets in the way. Perhaps we can clean up that area.
updated t0081-repo-pinning
- fixed a couple bugs with the tests
- made it a bit clearer (still a lot going on)
- the remaining tests are correct and highlight a problem with
pinning. Namely, that recursive pinning is buggy. At least:
towards the end of the test, $HASH_DIR4 and $HASH_FILE4 should
be pinned indirectly, but they're not. And thus get gc-ed out.
There may be other problems too.
cc @whyrusleeping
fix grep params for context deadline check
fix bugs in pin and pin tests
check for block local before checking recursive pin
- updated go-ctxgroup and goprocess
ctxgroup: AddChildGroup was changed to AddChild. Used in two files:
- p2p/net/mock/mock_net.go
- routing/dht/dht.go
- updated context from hg repo to git
prev. commit in hg was ad01a6fcc8a19d3a4478c836895ffe883bd2ceab. (context: make parentCancelCtx iterative)
represents commit 84f8955a887232b6308d79c68b8db44f64df455c in git repo
- updated context to master (b6fdb7d8a4ccefede406f8fe0f017fb58265054c)
Aaron Jacobs (2):
net/context: Don't accept a context in the DoSomethingSlow example.
context: Be clear that users must cancel the result of WithCancel.
Andrew Gerrand (1):
go.net: use golang.org/x/... import paths
Bryan C. Mills (1):
net/context: Don't leak goroutines in Done example.
Damien Neil (1):
context: fix removal of cancelled timer contexts from parent
David Symonds (2):
context: Fix WithValue example code.
net: add import comments.
Sameer Ajmani (1):
context: fix TestAllocs to account for ints in interfaces
Many times, a node will start up only to shut down immediately.
In these cases, reproviding is costly to both the node, and the
rest of the network. Also note: the probability of a node being
up another minute increases with uptime.
TODO: maybe this should be 5 * time.Minute
All the datastores used by pinners and so on should be mutex
wrapped. One issue with changing all of them from
ds.Datastore -> ds.ThreadSafeDatastore is that we wrap the
incoming ds.ThreadSafeDatastore with other datastores, which
do not implement the interface. Re-wrapping again causes
double locking. (which may be ok..., but...) any ideas?
