Maaz Zindani

Loading the index…

↑↓ move↵ openesc close4072 pages

Help for owners

My AWS database ran out of space and the app is down

Short answer

A full RDS database stops accepting writes, so the app errors. Storage can be increased by modifying the database, and RDS storage autoscaling grows it when free space is 10 percent or less of allocated storage, but only up to a maximum you set and no more than four changes in 24 hours. Storage cannot be reduced afterwards.

  • Written for business owners
  • Updated
  • By Maaz Zindani, Senior Software Engineer

Does this sound like your situation?

  • The app shows errors on saving, signing up or checking out, but pages still load.
  • The AWS console shows the database in a storage-full state.
  • My developer says the disk is full and asks about increasing storage.
  • It happened after a big import, a report or a log table grew.
  • I do not know how much space is used or what is taking it.

Why does this happen?

  1. Data and logs grew unnoticed

    Tables, indexes, temporary files and logs all use the same allocated storage. A table that stores every event or an unclean history can grow every day until nothing is left.

  2. Autoscaling was never switched on

    With storage autoscaling on, RDS raises storage when free space is 10 percent of allocated storage or less, the low-space condition lasts at least five minutes and fewer than four changes occurred in 24 hours. Without it, someone must raise storage by hand.

  3. Autoscaling hit its own ceiling

    You set a maximum storage threshold. AWS says autoscaling does not happen if the increment would exceed that maximum, and a very large data load can still leave the database storage-full for hours because further changes must wait for optimization and are limited to four per 24 hours.

  4. Storage cannot be shrunk

    After storage has been raised, RDS cannot reduce the allocated amount for the instance. AWS describes moving to a new smaller instance as the manual route. So an emergency raise is permanent.

  5. Some setups do not support autoscaling

    AWS lists exceptions, such as magnetic storage and a few older instance classes, where storage autoscaling cannot be used.

What can I do about it today?

  1. Confirm it is storageAsk your developer or open the RDS console and look at free storage space. Screenshot the graph and the database status.
  2. Do not delete data to free spaceDeleting rows in a hurry can lose records and often frees less space than expected. Raising storage is the safer first move.
  3. Take a snapshot if the database allows itA manual snapshot gives you a restore point before changes. If the console blocks it, tell your developer.
  4. Ask for a storage increase, with a sensible limitAWS lets you increase storage by modifying the instance. Ask for a clear new size and remember it cannot be lowered later.
  5. Turn on autoscaling with a maximumOnce calm, enable storage autoscaling and set a maximum that fits the business. A maximum protects you from runaway growth.
  6. Set an alert on free spaceA CloudWatch alarm on free storage warns before the next outage.

When should I bring someone in?

  • The database stays storage-full after you raised storage.
  • You do not know which table or log is eating space.
  • A big import or migration is planned.

How do I handle it?

  1. Stabilise firstSnapshot, raise storage by a measured amount, and confirm the app can write again.
  2. Find the growthI look at which tables, indexes and logs grew and on which dates.
  3. Fix the cause, not only the sizeOld log rows are archived, unneeded indexes are removed and retention rules are added, each with a backup first.
  4. Set limits and alarmsAutoscaling with a maximum, plus alarms at several levels of free space.
  5. Plan the next big loadLarge imports are sized in advance, since autoscaling alone may not keep up.

The price depends on what I find, so I quote in writing after a short discovery call. Tell me what is happening and I reply within 24 hours.

What do owners ask about this?

Will raising storage cause downtime?

AWS says changing autoscaling settings needs no reboot and no downtime. For manual storage changes follow the console's notice; I will check it before changing.

Can I reduce storage later to save money?

Not directly. AWS says allocated storage cannot be reduced after it has been allocated, and describes creating a new instance and migrating the data.

Is autoscaling enough on its own?

Not always. AWS says it cannot completely prevent storage-full situations for large data loads.

Sources

How was this page made?

I drafted this page with AI assistance from the official documents listed under Sources. Found a mistake? Tell me.

For your developer: the technical fixes

Next step

Want this off your desk?

Send me a few lines on where things stand. I reply within 24 hours.