Troubleshooting WinCC User Archive CSV Import Error 0x80044e20

1. Problem Context

On Siemens WinCC V7.x and WinCC Professional (TIA Portal) runtime installations, importing a CSV file into a User Archive (UA) component can fail with HRESULT 0x80044e20. The error is generated by the ImportCSV function inside CCUAImport.dll (the User Archive importer) and is logged in the WinCC diagnostics viewer (apdiag.exe) or the Windows Application log under source WinCC.

A representative failure from a redundant WinCC V7.5 SP2 server pair shows the exact format produced by the importer:

2023-01-04 14:58:48.182 [p=9112, t=0x22ec] ImportCSV - Archive: WALZ - Failed to insert dataset from CSV file '\\RMSRV2\WinCC_Project_RM_SERVER\ua\WALZ.csv'. Field headers = "ID";"walzcoll";"vzr30";"vtr40";"vge01";"vge02";"vge03";"vge04";"vge05";"vge06";"vge07";"vge08";"vge09";"vge10";"vge11";"vge12";"vge13";"vge14";"vge15";"vge16";"vge17";"Zge02";"Zge03";"Zge04";"Zge01";"Zge05";"Zge06";"Zge07";"Zge08";"Zge09";"Zge10";"Zge11";"Zge12";"Niveau";"Ziehtakt";"wadurch";"art";"vge114";"vge115";"vge116";"vge117";"vge118";"vdt141";"vge119";"vdt142";"vwl161";"vwt101";"vwt102";"vwt103";"vwt104";"Adt141d";"Age119d";"Sge114";"Sge115";"Sge116";"Sge117";"Sgo101";"Sgu101";"vege114";"vege115";"vege116";"vege117";"vedt141";"vege119";"vedt142";"vwdt141";"vwge119";"vwdt142";"vage119";"vadt142";"ILdt141";"ILge119";"ILdt142";"Gsge118";"groli";"gruli";"vge214";"vge215";"vge216";"vge217";"vge218";"vdt241";"vge219";"vdt242";"vwl261";"vwt201";"vwt202";"vwt203";"vwt204";"Adt241d";"Age219d";"Sge214";"Sge215...

2023-01-04 14:58:48.184 [p=9112, t=0x22ec] ImportCSV - Archive: WALZ - Failed to insert dataset from CSV file '\\RMSRV2\WinCC_Project_RM_SERVER\ua\WALZ.csv' line 1 (error=0x80044e20). Data set = 196;"2208 105";0,1817023;0,1815593;0,1792;0,2228;0,3043;0,3839;0,502;0,6576;0,8768;1,1431;1,5033;2,0359;2,6933;3,5673;4,71;0;0;0;0;3;3;3;3;3;3;5;5;5;5;5;5;100;90;10,5;"O";5,776;7,304;8,844;10,818;11,887;0;0;38,46;36,304;0,3799377;0,4;0,3999778;0,3999778;0;0;100;100;100;230;300;0;1;3;1;3;0;0;0;0;0;0;0;-1,5;0;0;190;6;350;250;5,831;7,221;8,813;10,72;12,052;0;0;38,87;37,006;0,3799377;0,4;0,3999778;0,3999778;0;0;100;100;100;230;300;0;1;3;1;3;0;0;0;0;0;0;0;0;-1,5;0;0;190;6;350;250;120;120;450;400;100;100;450;450;40;40;50;50;20;20;20;20;10;10;0;0;0;0;55;55;18;18;1,035;1,035;0,705;0,705;1,035;1,038;"";0;28.03.2016 0:21:32

2023-01-04 14:58:48.184 [p=9112, t=0x22ec] ImportCSV - Archive: WALZ - Import failed (error=0x80044e20).

Three diagnostic elements are visible:

• The archive name (WALZ) is resolved correctly by the importer.
• The CSV header row is parsed and reported in full so the engineer can count the declared columns.
• The first failing record is dumped together with the HRESULT. In the log line above the failure occurs already on line 1 of the data section, before any rows have been committed.

Symptoms on a redundant pair are typically:

• Import fails on the standby server (RMSRV2) but succeeds on the master.
• User Archives are not synchronized between master and standby after the failure.
• The WinCC Redundancy status in Redundancy Control shows the UA segment as FAULT instead of SYNC.

2. Decoding HRESULT 0x80044e20

The 32-bit HRESULT 0x80044E20 follows the standard Microsoft error structure:

Win32 error 20000 (0x4E20) is not a standard Windows kernel code; it is a SQL Server user-defined severity returned by the import procedure used by the WinCC User Archive database. The SQL Server sys.messages catalog ships the message as a custom User Archive error (range 50000-2147483647), so the value must be looked up against the running WinCC project's UA database.

To decode on the live system:

1. Open SQL Server Management Studio and connect to the WinCC instance WINCC (default) of the affected server.
2. Right-click the database that owns the archive (typically CC_UA_<project>_<server>, e.g. CC_UA_RM_SERVER_00).
3. Run SELECT * FROM sys.messages WHERE message_id = 20000;
4. If multiple language rows exist, filter with AND language_id = 1031 for German or 1033 for English.

The text returned in text is the human-readable cause; typical texts observed in production deployments are:

3. Root Cause Analysis: Field Header vs. UA Definition

The most common trigger of 0x80044e20 on redundant WinCC pairs is a divergence between the CSV header row and the field count defined in the User Archive configuration (UA Editor, part of WinCC Configuration Studio). When a column is added, renamed, deleted, or its data type is changed in the UA configuration but the CSV is not regenerated, the importer accepts the parsed header (WinCC logs it line-by-line), attempts the SQL INSERT with the wrong arity, and the database raises error 20000.

Inspect the failing case from the log:

1. The CSV header line lists 152 field names. Count them: ID, walzcoll, vzr30, vtr40, vge01...vwt204, Adt241d, Age219d, Sge214, Sge215 ...
2. The first failing data row contains: 196, "2208 105", 0,1817023, ... , "", 0, 28.03.2016 0:21:32
3. If the actual UA WALZ archive has 154 columns (two trailing columns added in a recent UA configuration change), every INSERT returns Msg 213 (column name ambiguity) or Msg 8152 (truncation/arithmetic) wrapped as 0x80044e20.

Three structural checks must pass before any CSV row will be accepted:

3.1 Detecting the Mismatch Programmatically

Connect to the SQL Server instance and dump the UA definition:

USE CC_UA_RM_SERVER_00;
GO
SELECT COLUMN_NAME, DATA_TYPE, CHARACTER_MAXIMUM_LENGTH, IS_NULLABLE, ORDINAL_POSITION
FROM INFORMATION_SCHEMA.COLUMNS
WHERE TABLE_NAME = 'UA#WALZ'
ORDER BY ORDINAL_POSITION;
GO

Compare the column count and order against the CSV header exported from the failing log. Any deviation in ORDINAL_POSITION is the root cause.

4. WinCC User Archive CSV Import Workflow

Understanding the import path clarifies why a server restart can resolve the failure. The import is processed through the following call stack on a redundant pair:

1. The WinCC UserArchiveControl or a C/VBS script calls UAImport or invokes the WinCC OLE DB provider on the UA#WALZ table.
2. The importer opens the CSV file (UNC path \\RMSRV2\WinCC_Project_RM_SERVER\ua\WALZ.csv) and parses the header line.
3. For every data row, an INSERT INTO UA#WALZ (col1, col2, ..., colN) VALUES (...) is generated.
4. If the active server is RMSRV2, the insert goes to the local SQL Server WINCC instance; if it is a remote master, it is forwarded via WinCC Redundancy using the RM service.
5. After commit, the change is added to the User Archive synchronization queue and replicated to the partner server.

A restart of the active server resets the following transient state:

• The in-process import handle held by the WinCC User Archive manager (CCUserArchive.exe).
• The SQL Server connection pool that may hold a stale prepared statement with the wrong column count from a previous UA configuration load.
• The redundancy replication backlog; on a clean restart the standby re-replicates the archive from the master.

This explains why the field engineer observed recovery after a service restart, even though the underlying schema mismatch was not corrected. The restart also rebuilt the UA schema cache from the SQL catalog, so the cached (and stale) column count from the previous UA edit was discarded.

5. CSV Format Requirements for User Archives

WinCC's ImportCSV function enforces the following rules. Failure to meet any of them is another common cause of 0x80044e20:

Reviewing the failing dataset against this table:

• Field separator is ; — correct.
• Decimal separator is , — correct for a German WinCC project.
• Date column shows 28.03.2016 0:21:32 — correct German format with day-first ordering.
• Empty text "" appears just before the date — acceptable for a nullable string column.
• The ID value of 196 must be checked for collision against the existing UA WALZ archive contents.

6. Step-by-Step Resolution Procedure

Step 1 — Verify UA Schema on Both Servers

1. Open WinCC Explorer on the master server and launch the User Archive editor.
2. Select the WALZ archive and note the column count and the order of column names exactly as they appear in the Properties panel.
3. Export the schema to a reference CSV using the editor's Export button — this CSV is guaranteed to match the UA definition.
4. Repeat on the standby server (RMSRV2). The two servers must report an identical column count and identical column names in identical order.

Step 2 — Export the Failing CSV Header for Comparison

1. Copy the failing file \\RMSRV2\WinCC_Project_RM_SERVER\ua\WALZ.csv to a local working directory.
2. Open it in Notepad++ or any text editor that exposes line numbers.
3. Count the semicolons in line 1 (the header line) and add 1 to obtain the column count.
4. Cross-check against the UA column count from Step 1.

Step 3 — Reconcile Any Mismatch

1. If column counts differ, regenerate the CSV from the UA editor's Export function so it is guaranteed to match.
2. If column names differ, replace the header line with the names exported by the UA editor.
3. If column order differs, reorder the CSV columns to match the UA ordinal positions.

Step 4 — Confirm Datatype Compatibility

1. For each UA column, note its type: INTEGER, FLOAT, DOUBLE, STRING, DATETIME.
2. Verify the corresponding CSV value passes parsing: FLOAT columns cannot contain ""; DATETIME columns cannot contain a numeric value.
3. Replace any non-conforming values with the empty representation expected by the locale (0 for numeric, "" for string).

Step 5 — Re-import

1. Stop the WinCC Runtime on the standby server first; on the master, open User Archive editor and select Import > CSV.
2. Browse to the corrected CSV file. Do not use a UNC path until the first successful run on a local path is verified.
3. Click Import. Watch the apdiag window for the ImportCSV - Archive: WALZ - prefix.
4. Confirm the success line: ImportCSV - Archive: WALZ - Import completed successfully.

Step 6 — Trigger Redundancy Sync

1. Open Redundancy Control on the master.
2. Right-click the WALZ archive in the segment list and select Force Sync.
3. Verify on the standby that the segment status changes from FAULT to SYNC within 30 seconds.
4. Confirm the row count on both servers matches via SELECT COUNT(*) FROM UA#WALZ from SSMS on each node.

7. Redundant Server Synchronization Recovery

If the importer fails on both servers and the archives are out of sync, perform the documented recovery sequence from the WinCC Information System under Redundancy > Failure of a Server:

1. On the partner that is still active, perform Project Duplicator > Save As with the option Configure and copy redundancy partner.
2. Shut down WinCC Runtime on both servers.
3. Restore the project on the failed server from the duplicate, ensuring the local SQL Server WINCC instance is in sync with the master.
4. Start WinCC Runtime on both servers in the order master first, standby second. The redundancy handshake re-establishes the RM channel and the UA segment.
5. After the handshake, the standby pulls the latest UA state via UA_REDUNDANCY_SYNC stored procedure; do not interrupt this for at least five minutes for archives over 100,000 rows.

8. Verification Steps

Use the following checks to confirm the import succeeded and redundancy is healthy:

9. Error Code Reference Table for User Archive Operations

The 0x80044E2x family is documented in the WinCC Information System under User Archives > Reference > Error Codes. Cross-reference the exact code against your installed help system (WINCC.chm) for the version-specific message text.

10. Preventive Measures

1. Single source of truth for schema. Always edit the UA schema only on the master server via WinCC Configuration Studio, then run Project Duplicator before touching the standby. Never edit a .UDD file directly in \<project>\ua\.
2. Always re-export the CSV after schema edits. Use the UA editor's Export function to obtain a template CSV; do not hand-craft the header line.
3. Version-tag the CSV. Append a comment line at the top of the CSV (lines beginning with #) noting the UA schema revision and export date. The importer will skip such lines if the SkipCommentLines registry value under HKLM\SOFTWARE\Siemens\WinCC\UA is set to 1.
4. Pre-validate before import. Use a SQL BULK INSERT into a staging table with the same column layout, then run a SELECT joining to the production UA to surface conversion issues before they reach ImportCSV.
5. Monitor redundancy state. Configure WinCC Alarm Logging to raise an alarm on @RM_FAIL_UA so a 0x80044E20 incident is surfaced immediately instead of being noticed days later during shift handover.
6. Service account privileges. The WinCC Runtime service account must have db_owner on the CC_UA_* database and read access on the source CSV share. Without db_owner, the SQL layer rejects inserts with Msg 229 wrapped in 0x80044E20.
7. Schedule bulk imports during maintenance windows. Imports of more than 50,000 rows block the UA replication queue and may stall live tag writes. Run them during planned downtime or use the WinCC UA's BulkInsert API for large loads.

11. Frequently Asked Questions