Hey, we have enabled the Diagnostics settings for our ADX cluster. Is it possible to query the .show commands, .show queries and other commands from log analytics? We want to create alerts whenever a query and/or command is failing and also want to use .show diagnostics to know the health of ADX cluster and other details.
@dgarg82, thanks for your question. Based on the survey results, we just started to work on this new integration (it is not available yet). With this new integration, you will be able to query “.show queries”, “.show commands” from Log Analytics, create alerts according to the query state (failed queries, for example), and create workbooks (visualizations). It would be great to hear more about your monitoring use-cases.
Thanks,Guy ReginianoAzure Data Explorer (Kusto)
@Guy_Reginiano - Thanks for the reply. Can you please let me know when by this feature would be available?
Also, can you please recommend what we can do to monitor the Kusto cluster Health? Is there any integration available with other system (Geneva/Jarvis) where we can check health of our kusto cluster?
@dgarg82, the ability to export the data of “.show queries”, “.show commands” to Log Analytics will be available in the following weeks. We will publish an article when it is available.
We are also working on new built-in monitoring experiences, including:
I am trying to build a service health score card for our program. Within in our program, we are using multiple azure services and Azure Data Explorer is the one we are using heavily. I want to understand that Is there a way where I can query the availability state of the kusto cluster using KSQL. I could see that there is a way to know the kusto service health using resource health (preview) feature.
Also, Is there a direct way where I can query the cluster memory usage and do you recommend any threshold which I should watch for any issue respect to performance of our cluster.
We have seen multiple issues in past where some of our customers were running repeated long queries which was exhausting memory thus impacting our cluster performance.
Thanks & Regards,
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.