Since key exchange can be started both ways simultaneously, and
these might not get finalised with UDP transport, we can't be
sure what encryption key will be used by the other side:
- the one derived in the key-exchange started by us,
- the one derived in the key-exchange started by the other node.
To alleviate this issue, we store two decryption keys in each session.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
For some reason pending packets were released in the wrong order
in simulation. A minimal delay between them fixes this.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
Nodes on single IP are limited in normal operation. This
would create issues for tests, so better turn off this
mechanism.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
If a handshake was already in progress, messages were dropped.
Instead of this, it is better to queue these and send as soon
as the handshake is finished and thus the encryption key is known.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
encodeMessagePacket checks for session and behaves differently
based on that. Exposing this difference in behavior.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
Sessions were limited to 256 for some reason.
For large scale operation it makes sense to allow more
open sessions (symmetric key sessions).
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
We do not need that many responses with FindNodeFast, since the
reposes can be ordered by distance
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
We really don't need this to be 4 seconds.
Later we should tune it better based on measurements
or estimates,
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
Make dht compile with new new (0.14+) nimble.
Compilation with Nimble <0.14 can't work, as
the many of the libraries on the dependency
tree are not version tagged. As we can't change
that, two paths remain:
1, use nimbus-build-system (i.e. git submodules)
2, use lock files introduced in nimble 0.14+
First was already working, this fixes second
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
It seems lockfile was generated wrong or not updated,
broken in 76bff752a2
setting it to version used in nimbus build system based build
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
initialize wait for response before sending request.
This is needed in cases where the response arrives before
moving to the next instruction, such as a directly connected
test.
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
run e.g. as
```
nim c -r -d:debug -d:chronicles_enabled=on -d:chronicles_log_level=TRACE -d:chronicles_sinks=textlines[nocolors,stdout] tests/dht/test_providers.nim >err
```
Signed-off-by: Csaba Kiraly <csaba.kiraly@gmail.com>
* track nim-libp2p's unstable branch and nim-bearssl's master branch
refactor accordingly: mainly switching from `import bearssl` to
`import bearssl/rand`, `BrHmacDrbgContext` to `HmacDrbgContext`, and related
changes
* fix ambiguous identifier
* nim 1.4 is deprecated
Co-authored-by: Michael Bradley, Jr <michaelsbradleyjr@gmail.com>
* initial providers manager implementation
* misc license update
* spelling
* rename provider tests
* add provider tests
* reexport provider manager
* only update provider record if it changed
* small bug in getProvByKey
* pass providers to both constructors
* fix records retrieval
* disable cache during tests
* avoid redefining data
* adding back tests with cache
* use `.errorOption`
* proper err msg unpacking
Csaba Kiraly
Dmitriy Ryajov