In RequestMessagesSync subscriber is listening to a feed where all whisper
events are posted. After we received event with a request hash - subscriber will
stop actively consuming messages from a feed, as a subscription channel will
get overflow and whole feed will get blocked.
Some events are posted to a feed before request is sent, so blocked feed results
in blocked sending.
Now we will unsubscribe after relevant event was received, and terminate subscriber
explicitly by timeout.
This commits adds support for postgres database.
Currently two fields are stored: the bloom filter and the topic.
Only the bloom filter is actually used to query, but potentially we will
use also the topic in the future, so easier to separate it now in order
to avoid a migration.
As part of a performance profiling of mailserver we noticed that most of
the resources on a query are spend decoding the whisper envelope.
This PR changes the way we store envelopes encoding the Topic into the
database key, so we can check that and we are able to publish the
envelope rawValue if it matches.
The change is backward compatible as only newly added envelopes will
have the new key, while old ones will have to be unmarshaled.
* Replace request ID when same request is restarted
* Remove unnecessary changes
* Execute all writes atomically only if request was processed succesfully
* Fix linter
* Fix shadowed errors
* Fix spelling
* Do not append same reference to a byte slice
* Notify users that envelope was discarded and retry sending it
* Update Gopkg files with released whisper version
* Forgot to remove signal after refactoring
* Split shhext.tracker into envelopes and mail monitors
* Send envelopes on every new attempt to deliver a message
* Re-send user payloads if previous envelopes weren't acknowledged
* Remove debug api across the codebase
Currently we only decrypt messages if received on the current bundle.
This changes the behavior so that messages can be decrypted if sent to
previous bundles as well, as otherwise is a bit restrictive
Currently PFS messages are decrypted and therefore modified before being
passed to the client. This make IDs computation difficult, as we pass
the whole object to the client and expect the object be passed back once
confirmed.
This changes the behavior allowing confirmation by ID, which is passed
to the client instead of the raw object.
This is a breaking change, but status-react is already forward
compatible.
This PR does a few things:
1) Add a call GetContactCode to check whether we have a bundle for a
given user.
2) Add a DH flag to the API (non-breaking change), for those messages
that we want to target all devices (contact-requests for example).
3) Fixes a few small issues with installations, namely if for example a
messages is sent without a bundle (currently not done by any client),
we still infer installation info, so that we can communicate securely
and making it truly optional.
We change the protocol to accomodate publishing multiple bundles, in
order to propagate bundles for group chats and have a way to extend it
further.
This commit re-introduces backward compatibility for direct messages,
to be removed once that is not an issue anymore.
`kdf_iter` parameter is reduced to 3200. This change is done because of
performance reasons, currently key derivation is too slow on some mobile
devices. The number of iterations before this commit is 64000, default
value in `sqlcipher` from version `3.0.0`.
fda4c68bb4/CHANGELOG.md (300---2013-11-05)
Implementation:
`sqlcipher_export` is used for migration, check out the link below
for details
https://www.zetetic.net/sqlcipher/sqlcipher-api/#sqlcipher_export
Change to support sending multiple bundles, as needed for group chats,
limit number of devices to 3 as already done in the UI and refresh
bundle daily.
This change allows to connect to the mail server that we were using before the app was restarted. Separate loop is listening for whisper events, and when we receive event that request was completed we will update time on a peer record.
Records are stored in leveldb. Body of the record is marshaled using json. At this point the only field is a timestamp when record was used.
This loop doesn't control connections, it only tracks what mail server we ended up using. It works asynchronously to connection management loop. Which tracks events that are related to connection state and expiry of the requests.
When app starts we look into the database and select the most recently used record. This record is added to connection management loop first. So if this server is available we will stick to using it. If we weren't able to connect to the same server in configured timeout (5s) we will try to connect to any other server from list of active servers.
closes: #1285
In previous change i used same filtering for a handler that processes confirmations.
But if we are connected with peers that do not support confirmations on whisper level,
whisper may send "Sent" event when envelope is written to a socket. This is our old behaviour.
So, this PR allows to use confirmations from a mail servers and ensure that transition between
multiple version of peers will be smooth.
This change implements connection manager that monitors 3 types of events:
1. update of the selected mail servers
2. disconnect from a mail server
3. errors for requesting mail history
When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster.
If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is.
If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above.
We will have two components that will rely on this logic:
1. requesting history
If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request.
2. confirmation from selected mail server
Confirmation from any selected mail server will bee used to send a feedback that envelope was sent.
I will add several extensions, but probably in separate PRs:
1. prioritize connection with mail server that was used before reboot
2. disconnect from mail servers if history request wasn't expired but failed.
3. wait some time in RequestsMessage RPC to establish connection with any mail server
Currently this feature is hidden, as certain changes will be necessary in status-react.
partially implements: https://github.com/status-im/status-go/issues/1285
We are preparing for the release of this to general public, so a few
things have been added:
1) Add versioning for bundles, and make refresh interval configurable
2) Move files to installationID so no metadata is leaked
3) Re-key using user password db
This commit updates geth to 1.8.17 and adds a possibility to enable metrics during compilation time.
The cascade of issues forced us to upgrade geth to 1.8.17 in order to allow enabling metrics during compilation time. 1.8.17 introduced `NodeID` refactoring and `enode` package which affected our peers pool and integration with Discovery V5.
This commit adds a list new table, installations, which is used to keep
track of which installation are active for a given identity key.
In general, we limit the number of installation that we keep
synchronized to 5, to avoid excessive usage of resources.
Any installation coming from our own identity, will have to be manually
enabled, otherwise we trust the other peer has correctly paired their
devices.
We use a timestamp to decide which installations to keep synchronized as
a logical clock would have make the creation of the bundle more
complicated, but this can always be converted to a logical clock at
later stages without breaking compatibility.
- Skipped keys
The purpose of limiting the number of skipped keys generated is to avoid a dos
attack whereby an attacker would send a large N, forcing the device to
compute all the keys between currentN..N .
Previously the logic for handling skipped keys was:
- If in the current receiving chain there are more than maxSkip keys,
throw an error
This is problematic as in long-lived session dropped/unreceived messages starts
piling up, eventually reaching the threshold (1000 dropped/unreceived
messages).
This logic has been changed to be more inline with signals spec, and now
it is:
- If N is > currentN + maxSkip, throw an error
The purpose of limiting the number of skipped keys stored is to avoid a dos
attack whereby an attacker would force us to store a large number of
keys, filling up our storage.
Previously the logic for handling old keys was:
- Once you have maxKeep ratchet steps, delete any key from
currentRatchet - maxKeep.
This, in combination with the maxSkip implementation, capped the number of stored keys to
maxSkip * maxKeep.
The logic has been changed to:
- Keep a maximum of MaxMessageKeysPerSession
and additionally we delete any key that has a sequence number <
currentSeqNum - maxKeep
- Version
We check now the version of the bundle so that when we get a bundle from
the same installationID with a higher version, we mark the previous
bundle as expired and use the new bundle the next time a message is sent
Adam Babik
Dmitry Shulyak
Andrea Maria Piana
Ivan Daniluk
Andrea Franz
Roman Volosovskyi :: Darkviolet Lightgreen Halcyon
Igor Mandrigin