# High schedule to start latency

**URL:** <https://community.temporal.io/t/high-schedule-to-start-latency/4964>\
**Category:** Community Support\
**Tags:** java-sdk, helm, cassandra, metrics\
**Created:** [June 14, 2022, 6:42pm UTC](https://community.temporal.io/t/high-schedule-to-start-latency/4964 "2022-06-14T18:42:43Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Parth\_Anand](https://sea2.discourse-cdn.com/flex016/user_avatar/community.temporal.io/parth_anand/32/2206_2.png) [@Parth\_Anand](https://community.temporal.io/u/Parth_Anand)\
**Post date:** [June 14, 2022, 6:42pm UTC](https://community.temporal.io/t/high-schedule-to-start-latency/4964/1 "2022-06-14T18:42:43Z")

</div>

**Deployment Architecture:**  
Worker processes: 14 pods with 13 GB memory/7 CPU cores  
Temporal frontend: 8 pods with 1 GB memory/1 CPU cores  
Temporal history: 25 pods with 2 GB memory/3 CPU cores  
Temporal matching: 17 pods with 2 GB memory/3 CPU cores  
Temporal workers: 20 pods with 2 GB memory/3 CPU cores.

We have configured workers with the following values:  
MAX\_CONCURRENT\_ACTIVITY\_POLLERS = 20  
MAX\_CONCURRENT\_WORKFLOW\_TASK\_POLLERS = 20  
MAX\_WORKFLOW\_THREAD\_COUNT = 5000  
MAX\_CONCURRENT\_ACTIVITY\_EXECUTION\_SIZE = 600  
WORKFLOW\_CACHE\_SIZE = 5000  
MAX\_CONCURRENT\_LOCAL\_ACTIVITY\_EXECUTION\_SIZE = 600  
MAX\_CONCURRENT\_WORKFLOW\_TASK\_EXECUTION\_SIZE = 600

**Problem:**  
There is a high schedule\_to\_start\_latency when we have around 200 workflows getting created per second.  
From metrics, we observe that a high number of task\_slots are available.

**What could be a probable reason behind this?**

---

<div class="post-metadata">

**Author:** ![Parth\_Anand](https://sea2.discourse-cdn.com/flex016/user_avatar/community.temporal.io/parth_anand/32/2206_2.png) [@Parth\_Anand](https://community.temporal.io/u/Parth_Anand)\
**Post date:** [June 14, 2022, 7:29pm UTC](https://community.temporal.io/t/high-schedule-to-start-latency/4964/2 "2022-06-14T19:29:03Z")

</div>

For everyone’s reference, some useful articles for the same:

for task queue partitions, can be set via “matching.numTaskqueueWritePartitions” and “matching.numTaskqueueReadPartitions” in dynamic config.  
Dynamic config knobs on server side: [read](https://github.com/temporalio/temporal/blob/0ff2418a38f0807f15283ba0574a4f120834b16e/common/dynamicconfig/constants.go#L149) [write](https://github.com/temporalio/temporal/blob/0ff2418a38f0807f15283ba0574a4f120834b16e/common/dynamicconfig/constants.go#L148)  
default is 4 and that should do the work, if not try setting to 8.