Closed WingsGuo8581 closed 3 years ago
could you also update the documentation to include that you need a new lua_shared_dict
? In this part: https://github.com/steve0511/resty-redis-cluster/blob/master/README.md#installation
Code looks good, thanks
could you also update the documentation to include that you need a new
lua_shared_dict
? In this part: https://github.com/steve0511/resty-redis-cluster/blob/master/README.md#installationCode looks good, thanks
Hi Tore! Thanks for your suggestion. Already updated the md. Pls check when u r available.
thanks for contributing.
Hi @steve0511
I'm trying to update this rock in my code so I can test this patch/fix. It seems it fix an issue we have under heavy load that matches the behaviour described here.
Would it be possible to create a release tag to include this (and other latest changes) after 1.02-4
release?
Thanks in advance.
Hi @steve0511
I'm trying to update this rock in my code so I can test this patch/fix. It seems it fix an issue we have under heavy load that matches the behaviour described here.
Would it be possible to create a release tag to include this (and other latest changes) after
1.02-4
release?Thanks in advance.
ok, create 1.03 based on latest.
In our project, the redis cluster consists of 32 masters and 32 slaves. resty-redis-cluster runs well most of the time. However, when there is failover and with high QPS, the fetch_slots method without lock will make the redis cluster crashed because of the high CPU cost of "cluster slots". Therefore, we add the refresh_slots method with lock to fix it and test in our production environment. We hope it will help other engineers if they encounter our situation. Thanks for your time!