LogoLogo
OverviewQuickstartsHow To GuidesReferenceArticlesSupport
Support
Support
  • Support
  • FAQs
    • Get Started with Upsolver
    • Basic Elements of Upsolver
    • Iceberg Cloud Storage Breakdown
    • Infrastructure
    • Cost Estimator
  • TROUBLESHOOTING
    • AWS Configuration
      • CloudFormation Stack Failed to Deploy
      • Private API Doesn't Start or Can't Connect
        • Elastic IPs Limit Reached
        • EC2 Spot Instance Not Running
        • DNS Cache
        • Security Group Not Open
    • Cluster
      • Compute Cluster Doesn't Start
      • Can't Connect to Apache Kafka Cluster
    • Jobs
      • Problem Ingesting Amazon S3 Data
      • Data Doesn't Appear in Athena Table
      • Exception When Querying Athena Table
  • ERROR MESSAGES
    • Error Messages
      • Cluster
        • UP10020 COMPUTE_CLUSTER is Missing
      • Jobs
        • UP10010 Missing ON Condition
        • UP10030 Entity Already Exists
        • UP10040 Entity Not Found
        • UP10050 Materialized View Illegal Column Expression
        • UP10060 Statement Parsing
        • UP10100 Cannot Select Records in an UNNEST Statement
        • UP20010 Source Data Not Found
        • UP20040 Could Not DROP Entity Used by a Job or Materialized View
      • Replication
        • UP20050 Reached PostgreSQL Replication Slots Limit
        • UP20051 PostgreSQL Replication is Disabled
      • Security
        • UP20020 No Access to Database
        • UP20030 No Permissions to assumeRole
        • UP20060 Unable to Connect
        • UP20061 Unable to Connect to a Private Network
Powered by GitBook
On this page
  • What is the testing and deployment process for new Upsolver versions?
  • What is a dry-run cluster?
  1. FAQs

Infrastructure

Learn about Upsolver's infrastructure and how this impacts your account.

Read the following FAQs to learn about the testing and deployment of new Upsolver versions and dry-run clusters.

What is the testing and deployment process for new Upsolver versions?

The continuous release cycle at Upsolver consists of constant improvements, up to twice a week, to the data ingestion mechanism and other infrastructure elements. Below is the process prior to the deployment of each software version to customers' clusters.

Stage
Environment
Details
Requirement

Feature-specific tests

Dev / QA

Based on the requirement of each feature or bug fix, tests are done to validate its performance and stability

All tests are completed successfully

Automated QA sanity tests

QA

Common platform usage scenarios are run to ensure no functionality has been affected by the code changes

All tests are completed successfully

Manual QA interface tests

QA

Main elements of the user interface are tested to ensure usability has not been affected by the code changes

All tests are completed successfully

2 Internal dry runs

Staging

The dry run mechanism runs both the existing and new versions on a small sample data set for each cluster, to verify data is handled correctly

Data processed and written by both versions is consistent and without differences

Customer clusters dry run

Production

The dry run mechanism runs both the existing and new versions on a small sample data set for each cluster, to verify data is handled correctly

Data processed and written by both versions is consistent and without differences

What is a dry-run cluster?

As part of testing a new version of Upsolver before deploying it to customers, dry-run clusters are created for a limited time to test the changes in safe, replicated environments.

Dry-run clusters mimic the setup and configuration of a production system, enabling our developers to safely check the new version on a replicated environment and discover any differences.

Occasionally, customers may notice instances in their accounts with a dry run suffix. These are used by Upsolver to improve the quality of our next release.

Details and schedule

Usually, dry-run clusters will run for 120 minutes, on Sunday and Wednesday afternoons, replicating all the clusters in an organization. However, in some cases, dry-run clusters run outside of these timeframes, and for shorter/longer times, depending on the testing requirements.

How does this affect you?

Customer data is not affected in any way, nor is the customer billed in the form of Upsolver credits for these instances. Customers are only billed for the uptime of the clusters as part of EC2 payment. Furthermore, dry-run clusters are created with only one EC2 instance, so they can mimic production tasks without raising high costs.