Problem Statement:
Many users leverage popular managed-database cloud providers such as Amazon Web Services (AWS), Google Cloud, Microsoft Azure, and more. Typically, there is an option for the DB’s to enable disk-space or storage auto-scaling of some kind.
In the context of Xray’s initial onboarding DBSync, however, this can sometimes cause a detrimental race condition, especially when a freshly installed Xray is unable to perform the initial onboarding DBSync process.
In this process, 200+ GB of CVE data would be added to the database at once; This can be faster than the auto-scaling mechanism activation, or otherwise the mechanism itself may trigger an interruption. Thus, the DB may forcibly close the DB connection abruptly and/or go into read-only mode. This leads to the following kinds of errors:
2026-07-21T18:29:36.102Z [jfxr ] [ERROR] [75c4f1516555ca52] [updates_progress:76 ]
[ui/updates/start ] Failed to increment progress of updates FATAL: terminating
connection due to administrator command (SQLSTATE 57P01)
2026-07-21T18:29:36.102Z [jfxr ] [ERROR] [75c4f1516555ca52] [updater:1085 ]
[ui/updates/start ] Failed to complete online PublicData update
--- at /go/src/jfrog.com/xray/xray/backend/backend/service/updates/updater.go:1002
(updater.handlePossiblyAbortedError) ---
Caused by: Failed to process online PublicData updates
--- at /go/src/jfrog.com/xray/xray/backend/backend/service/updates/updater.go:597
(updater.updateProcessCommon) ---
Caused by: Failed processing onboarding PublicData update
--- at
/go/src/jfrog.com/xray/xray/backend/backend/service/updates/executor/csv_updates_executor.go:200
(csvExecutor.ProcessUpdates) ---
Caused by: Failed rolling back the transaction
--- at /go/src/jfrog.com/xray/xray/backend/backend/dbaccess/dao/sql/sql.go:1069
(XrayDB.InvokeInPgxTransaction) ---
Caused by: conn closed\
2026-07-21T18:29:36.254Z [jfxr ] [ERROR] [75c4f1516555ca52] [updates_progress:98 ]
[ui/updates/start ] Failed to reset progress of updates ERROR: cannot execute UPDATE
in a read-only transaction (SQLSTATE 25006)
...
Caused by: ERROR: cannot execute SELECT FOR UPDATE in a read-only transaction (SQLSTATE 25006)
This also causes the DBSync status and UI to be stuck as “in-progress”, since the necessary status flags cannot be reset since the DB went into read-only mode. This leads to unnecessary waiting for a DBSync process which is not actually progressing.
Solution
Ensure that your DB used for Xray is provided an initial, dedicated 500+ GB of disk space (preferably with SSD and 3000 IOPS), as described in our docs here.
This will ensure the initial Xray DBSync will work as expected. This will also ensure that there is room for growth (i.e. users scanning their artifacts and entries being recorded in DB) as well time for monitoring before reaching the disk space limit (i.e. you notice 100 GB is left on the DB, and can plan accordingly to increase it).
While the auto-scaling option can still be used with Xray after this point, it is advised to plan for increases to the dedicated storage space periodically instead, to ensure Xray is not interrupted in any way via such a mechanism.