I want to enable storage autoscaling by the first time in a AWS RDS PostgreSQL instance.
Someone knows or have some documentation to clarify if this requires downtime? i can't found any articles or documentation about explicitly "enabling autoscaling by the first time"
Thanks in advance.
From Amazon RDS now supports Storage Auto Scaling:
RDS Storage Auto Scaling automatically scales storage capacity in response to growing database workloads, with zero downtime.
The message you highlight suggests that downtime would only be caused by other changes in the "pending modifications queue" (eg a requested change of instance type).
Related
I have created one DocumentDB cluster in AWS with two instance running in it, but I need to know the exact storage which will be used for storing the data and also how AWS charge for one cluster.
When you provision an Amazon DocumentDB cluster, you don’t need to specify how much storage or I/Os you need for your cluster. Amazon DocumentDB uses a unique storage system that automatically scales from 10 GB up to 64 TB of data per cluster in 10 GB increments.
Storage is at the cluster level, which means all your instances share the storage. You can view how much storage are you using my monitoring the VolumeBytesUsed metric in the Monitoring tab of your Amazon DocumentDB console. Storage in DocumentDB is priced as low as $0.02 per GB/month (prices may vary across AWS regions). Details here - https://aws.amazon.com/documentdb/pricing/. To how much you are paying for storage, you can also go to the AWS Billing console and view the details of your DocumentDB bill
It is my understanding that Redshift is built for performance but not for Availability.
The documentation https://aws.amazon.com/redshift/faqs/
suggest that once any node is down, the whole cluster is down until the node is restored. In the case of the AZ failure, you have no luck at all.
This post suggests having a double cluster
https://aws.amazon.com/blogs/big-data/building-multi-az-or-multi-region-amazon-redshift-clusters/
however, it is not clear to me how do you replicate Looker's PDT tables to support instant failover via Route 53 to the standby cluster?
Just curious about what people do to address the HA issue on Redshift?
from the Q&A regarding high availability in case an AZ disruption scenario:
"If your Amazon Redshift data warehouse cluster's Availability Zone becomes unavailable, Amazon Redshift will automatically move your cluster to another AWS Availability Zone (AZ) without any data loss or application changes. To activate this, you must enable the relocation capability in your cluster configuration settings." https://aws.amazon.com/redshift/faqs/?nc1=h_ls
Redshift now supports multi-AZ deployments: https://aws.amazon.com/redshift/reliability
I am looking at either setting up Aurora Postgresql or RDS Postgresql instance in AWS.
I would like the db instance to be running in 2 different regions and would like real time replication to be set up. I would also like no downtime for rehydration / patching etc.
Based on what I have read / discussed with colleagues so far , I am under the impression that Aurora Postgresql is the option to choose because RDS needs few minutes of downtime for rehydration and Aurora supports realtime replication of db instance across different regions.
Is my understanding correct and are there any other factors that I should be aware of?
No RDS product supports "real-time" replication across regions. Cross-region replication is always asynchronous.
You can expect to see a higher level of lag time for any Read Replica that is in a different AWS Region than the source instance, due to the longer network channels between regional data centers.
https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/USER_ReadRepl.html#USER_ReadRepl.XRgn
Additionally, cross-region replicas for Aurora/Postgres are not yet available.
Cross-region replicas are only available for Aurora/MySQL... but a cross-region replica is not for zero downtime or failover, anyway -- it's only for geo/latency-based read scale-out or disaster recovery, because once you promote the replica, the original master has to be abandoned, because replication is one-way.
If, when you said "region," you were actually referring to availability zones, then that is much more straightforward, since the backing store of Aurora instances is replicated across 3 availability zones within the region, and replication is synchronous. All replicas in a single region can be synchronous, even in different AZs, since they all share the same replicated storage.
Is there some kind of native Postgres tool they use, or is it a custom one? Are the replicas always in sync or do they drift apart from time to time?
With Multi-AZ RDS replication is synchronous. And since AWS like to be in full control of their software, it’s most likely a customised replication (but I couldn’t tell you for sure).
We created a second gen. Google Cloud SQL database with a failover replica and we've been looking into upgrading the CPU and memory to increase capacity. I've been able to change the tier of the first server by updating and saving the instance; however; when it comes to updating the failover the option to change tiers is greyed out
Is this intentional with the expectation that you delete and recreate the failover with a higher CPU/memory tier or am I missing something?