If your Cloud Platform application on Cloud Classic infrastructure uses its set quantity of allocated disk storage or exceeds its total number of allotted inodes, your application’s file system or database stop working, causing your application to stop working. You can examine your disk storage availability from the command line.
If you are approaching your storage limit, Acquia recommends that you increase the amount of storage available. Approaching the storage limit puts all applications and environments at risk for immediate downtime. For example, if your volume is 95% full due to your site's files area, you must take immediate steps as recommended. In certain cases, Acquia Support might intervene to temporarily free up space by removing temporary files or arranging for storage resizing. For long term resolution, consider scheduling an upsize before reaching full capacity.
Here are a few recommended actions that you can take when your application approaches the storage limit:
- Remove any on-demand database backups you do not need.
- Remove large, third-party, temporary, or outdated files that your applications do not require. Typically, such files are stored in
sites/default/files,sites/all/files, etc. - Remove unwanted table rows or truncate large tables containing data you do not need.
- Clean up temporary files that are no longer necessary.
- Check your code repository for any large files and remove them using tools like BFG Repo-Cleaner. In addition, avoid saving files to root directories like
/tmp; instead, use directories like/mnt/for backups. - Ensure that you drop all sections below 95% total usage to prevent further notifications about volume risk.
Additional disk storage may have an additional cost; for details. For more information, visit About Cloud Platform Billing: Disk storage. In addition, increasing your disk storage allocation makes all sites on your infrastructure unavailable for 30 minutes or even longer (depending on the size of the disk, how many files and how much data you have) while the larger disk volume is provisioned and your data is transferred from the old disk to the new one. As a rough estimate, the process takes approximately one minute per GB of data on your existing disk, plus approximately five minutes of additional setup time. For information about resizizng your application’s storage allocation, visit Managing Cloud Platform infrastructure.
Inode usage in your filesystem
In Unix-style file systems, inodes are data structures that describe files and directories, but do not contain the actual data of the file. The maximum number of inodes—thus the maximum number of files—for your infrastructure is set when Acquia provisions your infrastructure.
In most cases, infrastructure runs out of file space due to a small number of files that consume all available storage space on the disk. However, it is possible to create a large enough number of tiny files that exceed the total number of inodes your infrastructure can support. If this situation occurs, your application will experience data loss and potentially website downtime until the total number of inodes is reduced.
If Acquia Support notifies you of a disk-full issue, you should review both the space used in your filesystem and your inode usage.
Acquia may also provide you with reports similar to the following example to help you determine where to remove files:
Inode usage on [web-0000]:
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/xvda1 640K 222K 419K 35% /
tmpfs 471K 1 471K 1% /dev/shm
/dev/xvdb 979K 320K 659K 33% /mnt- Inodes: the total number of inodes available in this region of the filesystem
- IUsed: the number of inodes in use in this area
- IFree: the number of additional inodes needed to completely fill this region of the filesystem
- IUse%: the percentage of available inodes currently in use
- Mounted on: the path to this region of the filesystem