Duplicate Cache PDFs Created When Loading Document from API or Query

Problem reported by Kavindu Shavinda - 9/25/2025 at 2:55 AM

Resolved

Hi,

I am using GleamTech DocumentUltimate (7.6.6).

I have implemented a custom Document Viewer setup where a file can be loaded from a static file path, an API, or a SQL query.

When I load a document from a static file path, the file is displayed correctly and caching works as expected.
However, when I load the document from an API or a SQL query, I notice that two cache PDFs are created in the cache location. These duplicate cache files are not created when using the static file path option.

Is this the expected behavior when using API or query-based document loading? If not, how can I avoid duplicate cache PDFs from being generated?

Thank You.

3 Replies

Reply to Thread

Cem Alacayir Replied

9/25/2025 at 11:46 AM

Employee Post

This looks like the result of keepVariations or CacheKeepVariations

keepVariations Boolean (Optional): Whether to keep variations of document conversion results. The default is false.
For instance when you change watermark options, both Pdf and Xpz outputs are regenerated. By default the Pdf and Xpz outputs are replaced in cache when watermarks are changed. If you set this property to true, the Pdf and Xpz outputs will not be replaced but all variations will be kept in cache instead. This can be useful especially when you want different outputs for different users. So this way, for instance you can have different watermarked variations for each different user.

So if you are creating DocumentCache in your API server directly, maybe you are passing a different value for this parameter (default value is actually false for DocumentCache or DocumentUltimateWebConfiguration).

Kavindu Shavinda Replied

9/26/2025 at 2:39 AM

In my case, I have not configured those properties anywhere, so I don’t think they are causing the duplicate cache files. I suspect it might be related to my custom handler implementation.

I recently upgraded DocumentUltimate from v6.2.1 to v7.6.6. In the older version (6.2.1), my custom handler implementation worked fine and I did not see duplicate cache PDFs being generated. After upgrading to 7.6.6, I had to refactor some code because a few internal classes changed, and now I’m noticing this caching behavior.

public override string File
        {
            get;
            set;
        }

        public override bool CanGetInfo => true;

        public override bool CanOpenRead => true;

        public override bool CanOpenWrite => false;

        public override bool CanSerialize => false;

        protected override FileProviderInfo DoGetInfo()
        {
            byte[] fileContent = new byte[0];
            DateTime lastModified = DateTime.UtcNow;

            string sql = string.Format(selectCommand, parentKey);
            var dataSet = DataServiceBase.ExecuteDataset(connectionString, CommandType.Text, sql);
            if (dataSet.Tables.Count > 0 && dataSet.Tables[0].Rows.Count > 0)
            {
                if (isFileNameMapped)
                {
                    this.File = dataSet.Tables[0].Rows[0][fileName].ToString();
                }
                 
                if (mustBase64Decode)
                {
                    fileContent = Convert.FromBase64String(dataSet.Tables[0].Rows[0][fileContentsField].ToString());
                }
                else
                {
                    fileContent = (byte[])dataSet.Tables[0].Rows[0][fileContentsField];
                }
            }

        return new FileProviderInfo(this.File, lastModified, fileContent?.LongLength ?? 0);
        }


        protected override Stream DoOpenRead()
        {
            byte[] fileContent = new byte[0];

            string sql = string.Format(selectCommand, parentKey);
            var dataSet = DataServiceBase.ExecuteDataset(connectionString, CommandType.Text, sql);
            if (dataSet.Tables.Count > 0 && dataSet.Tables[0].Rows.Count > 0)
            {                   
                if (mustBase64Decode)
                {
                    fileContent = Convert.FromBase64String(dataSet.Tables[0].Rows[0][fileContentsField].ToString());
                }
                else
                {
                    fileContent = (byte[])dataSet.Tables[0].Rows[0][fileContentsField];
                }
            }

            return new MemoryStream(fileContent);
        }

Cem Alacayir Replied

10/20/2025 at 1:56 AM

Employee Post Marked As Resolution

You need to set this property to true in your FileProvider:

public override bool CanSerialize => true;

From the docs:

Implement this property to indicate if a file with this provider can be serialized across processes and machines.
For instance, a file on disk can be serialized because you only need path information to restore it but a
file in memory is volatile, i.e. it cannot be restored on other machines.
This property is mainly used by caching mechanisms, for deciding if a local copy of the file should be made.
If the file is serializable, there will be no need for a local copy.

So you see the additional PDF in the cache because it's a local copy as your FileProvider thinks it will not be able to get the same file again when required.

Back to Community Threads

Reply to Thread

Enter the verification text