By using AWS re:Post, you agree to the AWS re:Post Terms of Use
AWS re:Postre:Post

What are recommended alarms in Cloudwatch for Elasticache Serverless for Redis?

0

I would like to query my Elasticache Serverless instance for Redis by sensible metrics and alarms.

I found several alarms here. However, they seem to be unrelated to the serverless regime.

i am also interested to see when the Cache approaches the size limit or the "ECPU per second" limit. However, there are also other limits we might want to take into consideration for ensuring reliability.

Hence, could you please me know how to follow best practises here?

Topics
ServerlessDatabaseManagement & Governance
Tags
ServerlessAmazon ElastiCache for RedisAmazon ElastiCacheAmazon CloudWatch
Language
English
rePost-User-3708451
asked 16 days ago191 views
1 Answer
0

Hi

Here are some recommended CloudWatch alarms for your serverless instance:

Resource Utilization:

  • CPU, Memory Utilization
  • Evictions - High eviction rates indicate your workload might not fit in the current cache size.
  • CurrConnections - The number of client connections to your cache.
  • Latancy

You can check this link for more information of the metrics https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/serverless-metrics.html

ECPU per second limit: I think this metric isn't directly available in CloudWatch for serverless instances, but you can monitor overall CPU utilization as a proxy.

Reference Links from Official sources:

Monitoring best practices with Amazon ElastiCache for Redis using Amazon CloudWatch: https://aws.amazon.com/blogs/database/monitoring-best-practices-with-amazon-elasticache-for-redis-using-amazon-cloudwatch/

Which Metrics Should I Monitor?: https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/CacheMetrics.Redis.html

Serverless Redis: https://aws.amazon.com/blogs/aws/amazon-elasticache-serverless-for-redis-and-memcached-now-generally-available/

profile picture
EXPERT
GK
answered 16 days ago
profile pictureAWS
EXPERT
AWS-User-alantam
reviewed 16 days ago
  • rePost-User-3708451
    16 days ago

    Many thanks for your input. Could you tell me how CPUUtilization serves as a proxy for ECPU per second? Also, how exactly should I configure the alarms for the metrics you mention? Shouldn't there be an alarm on bytes consumed in comparison to the maximum number of memory?

  • GK EXPERT
    15 days ago

    ECPU (ElastiCache Processing Unit) per second is a metric specific to provisioned ElastiCache instances. It reflects the actual processing power allocated to your cache. Higher CPU utilization typically indicates the cache is working harder. This could be due to increased workload or inefficient operations. If CPU utilization stays consistently high (e.g., above 70-80%), it suggests your workload might be exceeding the allocated resources. This could lead to throttling, impacting performance. Therefore, monitoring CPUUtilization helps identify potential issues that might be related to ECPU limitations in a serverless environment.

  • rePost-User-3708451
    15 days ago

    Many thanks for getting back to me, however, not only don't I find something like "ECPU per second", I also do not find "CPUUtilization" for the serverless Redis cluster. Do you have an idea how to resolve that?

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions

Relevant content