none
Memory usage shot up RRS feed

  • Question

  • Hello,

    Today I noticed my memory usage shot up from around 100 bytes to nearly 45MB out of nowhere. I only have 3 keys with a very small amount of data stored against each. I also noticed my server load now keeps spiking to 1%.

    This seemed to coincide with me upgrading from the Basic C0 tier to the Standard C0 tier, could that be the cause? Or could it be Microsoft maintenance? I can't think of anything else that would cause my memory usage to jump up so significantly despite having close to no data.

    This is my configuration:

                        "maxclients": "256",
                        "maxmemory-reserved": "0",
                        "maxfragmentationmemory-reserved": "0",
                        "maxmemory-policy": "volatile-lru",
                        "maxmemory-delta": "0"

    • Edited by jbd312 Tuesday, October 1, 2019 4:17 PM
    Tuesday, October 1, 2019 3:15 PM

Answers

  • Hi jbd312,

    If you have the Redis console set-up, you can run a series of commands to further investigate this issue where this information is found: https://redis.io/commands#

    To set-up the console: How to use the Redis command-line tool with Azure Cache for Redis (link)

    The red boxes are commands you can use to initially look at your deployment and the red arrows can be leveraged to further investigate what you see. 

    The one main difference between Basic and Standard is that Standard offers a 99.99% SLA which means the nodes are replicated. There are additional background processes to account for this and that is what you likely are seeing but the above commands will help you dig a little deeper:

    Standard

    "By default, a replicated cache is a two-node primary/secondary configuration. We manage automatic replication between the nodes to offer a high-availability SLA. The prices below show the cache price per node."

    Aside from the Redis commands, there is also monitoring functionality that allows for proactive alerting that you can configure once you have a baseline of your deployed instance established.

    Please let us know if you have additional questions.

    Regards,

    Mike

    Wednesday, October 2, 2019 3:56 PM
    Moderator

All replies

  • Hi jbd312,

    If you have the Redis console set-up, you can run a series of commands to further investigate this issue where this information is found: https://redis.io/commands#

    To set-up the console: How to use the Redis command-line tool with Azure Cache for Redis (link)

    The red boxes are commands you can use to initially look at your deployment and the red arrows can be leveraged to further investigate what you see. 

    The one main difference between Basic and Standard is that Standard offers a 99.99% SLA which means the nodes are replicated. There are additional background processes to account for this and that is what you likely are seeing but the above commands will help you dig a little deeper:

    Standard

    "By default, a replicated cache is a two-node primary/secondary configuration. We manage automatic replication between the nodes to offer a high-availability SLA. The prices below show the cache price per node."

    Aside from the Redis commands, there is also monitoring functionality that allows for proactive alerting that you can configure once you have a baseline of your deployed instance established.

    Please let us know if you have additional questions.

    Regards,

    Mike

    Wednesday, October 2, 2019 3:56 PM
    Moderator
  • Thanks for your reply, seems like the replication was the cause of the memory increase.
    Monday, October 7, 2019 8:13 AM