Reviews from AWS Marketplace
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
External reviews are not included in the AWS star rating for the product.
Does not scale
What do you like best about the product?
Flexibility -- you can add database columns on-the-fly at run-time.
UDP or TCP -- go with UDP for efficiency or TCP for guaranteed delivery.
Easy time series data -- that's what it's made for.
UDP or TCP -- go with UDP for efficiency or TCP for guaranteed delivery.
Easy time series data -- that's what it's made for.
What do you dislike about the product?
Poor performance -- queries to check data would often timeout and fail.
Limited SQL -- very limited SQL syntax was frustrating -- queries from Internet could not be copied and pasted.
Unreliable -- could not handle all of are data, so we throttled and threw away most of the data, defeating the purpose.
Too flexible -- too easy to accidentally create duplicate columns. It's like going back to BASIC programming days where variables don't have to be declared.
Limited SQL -- very limited SQL syntax was frustrating -- queries from Internet could not be copied and pasted.
Unreliable -- could not handle all of are data, so we throttled and threw away most of the data, defeating the purpose.
Too flexible -- too easy to accidentally create duplicate columns. It's like going back to BASIC programming days where variables don't have to be declared.
What problems is the product solving and how is that benefiting you?
We used it for logging information from mobile apps. Arguably it is not a logging product, but a series of log messages is a time series as far as I am concerned.
The logs did help us troubleshoot problems -- when we could get the SQL syntax to work and not time out.
The logs did help us troubleshoot problems -- when we could get the SQL syntax to work and not time out.
Recommendations to others considering the product:
Be sure it is fit for your purpose and can handle enough data before you commit to the tool
- Leave a Comment |
- Mark review as helpful
Influx DB for operational monitoring
What do you like best about the product?
Easy to setup and customize, no performance issue so far (monitoring millions of events daily).
Has proven to scale without having to resort to technical wizardry.
Has proven to scale without having to resort to technical wizardry.
What do you dislike about the product?
Nothing. INFLUX DB does what it is supposed to do.
What problems is the product solving and how is that benefiting you?
Operational monitoring. Facilitates an agile introduction of additional monitoring, independently of the actual source, format and protocols of the information source.
A good choice for timeseries data
What do you like best about the product?
The most thing I liked was the ability to store very big timeseries data in very efficient way. In my case, we stored relatively big data with single node which uses casual compute and storage resources
What do you dislike about the product?
Although influxdb core is open source and free, using it in production and at scale will require a very expensive license. For example, clusters are only available in InfluxEnterprise which was costy for us in early development phase.
What problems is the product solving and how is that benefiting you?
We use it to store our infrastructure health and usage metrics. The most benefit in this use-case was the good support for tool-chains.
Recommendations to others considering the product:
Influxdb is designed to store timeseries data, not anything else. If you data nature is a timeseris that can be displayed in line graphs so go for influxdb, otherwise it may not be the best fit. Influxdb don't have *good* support for strings, and no support for booleans nor nested documents. So double check your data is a timeseries, and you will require lots of time-based queries.
Great time series database with an easy to usy to use syntax
What do you like best about the product?
The syntax is intuitive and easy to learn, as it is very similar to standard SQL queries. Learning the different aggregations and advanced features takes some time, but afterwards they are very easy to use. Downsampling data points to a lower resolution and deletion of older data can be automated very easily and reliably. The data compression is amazing (among the best compared to other time series databases), with a single data point with timestamp generally being less than 10 bytes in total for long-term storage and they are continuously updating the storage engine for even better compression. Integration with other applications in the TICK stack is easy and seamless and allows for even more advanced manipulation of the data. Documentation is
What do you dislike about the product?
Debugging the application to determine optimal configuration settings for performance/load is tricky and the documentation regarding this is lacking.
What problems is the product solving and how is that benefiting you?
We use InfluxDB as our main time series database, holding data written from 100s of sources. This data can then be manipulated and requested back for business intelligence/machine optimization.
Recommendations to others considering the product:
Consider your database design carefully to minimize cardinality, since it can affect the performance of the system in a major way if this is designed improperly.
Amazing
What do you like best about the product?
How easy it is to configure and get running
What do you dislike about the product?
The lack of joins across measurements can be annoying at times
What problems is the product solving and how is that benefiting you?
Collecting system measurements from various network devices
Recommendations to others considering the product:
Understand what your schema should be like and set up your tags and values properly
My most favourite time series database ever!
What do you like best about the product?
Easy deployment, single binary unbeatable. Compared with Graphite which was state of the art when I started using InfluxDB
What do you dislike about the product?
Pseudo-SQL syntax sometimes tricky because of perceived faux familiarity. Also in it's usual habitat (docker container) the REPL has no autocomplete or tab completion which can make editing queries tricky. Fortunately this doesn't have to be done much, as most people with use Grafana in front of it most of the time.
What problems is the product solving and how is that benefiting you?
Metrics data in general (alerting via Grafana) and also as principle data store of time series data which is shared with customers via our web application. In that regard, for some of our data InfluxDB is a primary data store for business critical operations.
Recommendations to others considering the product:
Look beyond the fact that the vendor wants you to use their whole stack. Their TICK stack has strenghts and weaknesses. InfluxDB coupled with Grafana works wonderfully, and Telegraf is a good "metrics collector" too.
I have been fairly disappointed in InfluxDB
What do you like best about the product?
How easy it is to pump data into InfluxDB, query it, and downsample it.
The query language is great, intuituve, and well thought out.
It is fairly easy to install, configure, and get working out of the box.
The query language is great, intuituve, and well thought out.
It is fairly easy to install, configure, and get working out of the box.
What do you dislike about the product?
Performance. I have very little data (100k points) on an Amazon EC2 t2.medium instance (below recommendations, but I have very little data) and when I query for data it crushes the CPU. And is very slow to return results.
What problems is the product solving and how is that benefiting you?
Tracking site wide metrics via timestamped events. It allows us to track a variety of metrics easily and query on them as we think of new ways to massage the data.
THE database for time series
What do you like best about the product?
This is the database to go for if you have time series to store. Optimized for that functionality with no bling bling features that have no relevance. I like that you can communicate with it with HTTP directly. Very useful for web applications. Also, the tags system is very useful to annotate your data without adding redundancy in your database. I like also the storage engine that do not take so much disk space to store a big amount of data, which is very useful for use cases like Internet of Things where you have to store multiple values per points in very small ranges of times (every 30s or less).
What do you dislike about the product?
The documentation states it as a CRud database, e.g. it is developed to Create and Read from it, but not really for Update or Delete, which is a logical thing to do in a time series database. But sometimes you want really to delete or update some points, which is not as simple as the word update/delete seems.
What problems is the product solving and how is that benefiting you?
The problem we solved with Influx is to never have to deal with time. It is automatically managed in a very cool way. You just throw the values at it and its done. This allows the developer to focus more in their server logic and have more time and effort to fix that without worrying about the data he is working with.
Recommendations to others considering the product:
You just need to look at their roadmap, as Influx is still in development.
showing 91 - 98