zip all attachments to a record, add the zip as an attachment and delete the original attachments?

[stevemac]

We are approaching the 4TB limit on our instances and do not want to pay for an increase.  The biggest table is sys_attachment_doc.  As such we are looking at options to reduce the size.  In consideration are

1. Delete attachments for some older records
2. Extract attachments and archive on our storage, delete the attachments in ServiceNow
3. for some records, compress (zip) their attachment(s), add the compressed file as an attachment and delete the original attachments

For option 3 - does anyone have insight on how to do this via script from within ServiceNow - specifically creating the compressed file that contains the attachments?

Overall I'd think the flow would be

• Identity the records that can have their attachments compressed.
• Check the attachments.  For non-compressed files - add them to a zip file
• Attach the zip file to the record
• delete any attachments that are not compressed

thanks,

Steve

[-O- · Accepted solution]

Supposedly one can do it using Flows: Zip and unzip attachments in #servicenow. Courtesy of @Uncle Rob. But it "needs" Tokyo.

[Uncle Rob]

Thanks for the shout out!  Now that Utah's out, Tokyo is going to be expected for most customers.

[Uncle Rob]

Talk to your legal and compliance teams about data retention.  In many cases orgs don't want to keep data more than X years old.  You can use archiving and deletion rules at that stage.

Also I once had the dubious honor of having the largest SN database.
We had a process group that was uploading Excel files to tickets every day.
Somehow they were being saved in preposterously large file sizes (gigabytes per file)
So we got that team to fix the problem saving their excel files and the problem went away.
You could look for some common causes to the largest file size problems.

[stevemac]

Excel - known to cause some headaches.  We have the the ability to paste rows of config item (names) into a field in a change request and a background process updates the affected CI table with them.  Had issues with users pasting in 1M mostly empty rows from excel which may have caused some performance issues.  We now strip out empty rows before saving / processing.