Related to #798. This gem has been causing outages lately as we scale. Right now, I'm still seeing slow Lua scripts when deleting items from large sets, e.g. when a spike of requests come in and jobs are queued up for each request, or when a lot of scheduled jobs are pushed to the queue. I tried to fix this upstream, but there's more issues and I don't have time to debug.
Follow up to fc1b739dab603c45d6fe00fe99e77138a69c30d6 (i.e. the fix for the big outage), which removed all mission-critical use of the gem. If and when we need job uniqueness, we can reach for Sidekiq Ent, which has timing guarantees.
Related to #798. This gem has been causing outages lately as we scale. Right now, I'm still seeing slow Lua scripts when deleting items from large sets, e.g. when a spike of requests come in and jobs are queued up for each request, or when a lot of scheduled jobs are pushed to the queue. I tried to fix this upstream, but there's more issues and I don't have time to debug.
Follow up to fc1b739dab603c45d6fe00fe99e77138a69c30d6 (i.e. the fix for the big outage), which removed all mission-critical use of the gem. If and when we need job uniqueness, we can reach for Sidekiq Ent, which has timing guarantees.