-
Notifications
You must be signed in to change notification settings - Fork 335
[WIP] Use asyncio protocol #1287
base: master
Are you sure you want to change the base?
Conversation
This pull request introduces 6 alerts and fixes 1 when merging fbedfbe into 659a14d - view on LGTM.com new alerts:
fixed alerts:
|
Codecov Report
@@ Coverage Diff @@
## master #1287 +/- ##
=======================================
Coverage 90.75% 90.75%
=======================================
Files 21 21
Lines 6947 6947
Branches 889 889
=======================================
Hits 6305 6305
Misses 469 469
Partials 173 173
Flags with carried forward coverage won't be shown. Click here to find out more. Continue to review full report at Codecov.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry, I'm now in a different team that's not working with aioredis, so I'm not available for reviewing.
thanks for all the help along the way @bmerry ! |
General update here - Anything involving a MONITOR or PubSub SUBSCRIBE is not currently compatible. This is because both of these commands create an open channel which streams data as it is received, but the current Protocol implementation attaches all received data into a single future which is associated to the original packed command. We'll need an alternative implementation which allows these commands to listen on the stream decoupled from the original sent command. |
This pull request introduces 2 alerts and fixes 1 when merging 834d431 into 659a14d - view on LGTM.com new alerts:
fixed alerts:
|
Hi guys. Any update? Performance degradation is blocks an update to 2.* version in our projects =( |
@matemax all updates are going to redis/redis-py. This repo at some point will convert back to its 1.3.1 version |
This supersedes #1216 and is an attempt to address the latency issues in #1208 -
Mainly I've gone through and added a Python parser which matches the hiredis implementation as closely as possible. I've also made a few adjustments to the RedisProtocol as introduced by @m-novikov in #1216.
I spent all day yesterday trying to get perfomance for this port to match something close to what @artesby & @m-novikov reported in their benchmarks in #1208, but I have a suspicion it's not possible.
@m-novikov and @artesby I'd be much obliged if you could test this implementation to see if we're actually going the right direction with this change. It's not yet ready for the real world and there are a few failing tests, but I don't want to keep barking up this tree if it won't get us where we need to go.