Delete orphaned assets

I was asked this question: we have about 3TB of assets, any way to clean it up.

These days, storage is cheap, but still not free. and big storage means you need space for back up. and with that, bandwidth and time.

Is there away to clean up things you no longer need?

Yes!

Optimizely Content already has a scheduled job named Remove Abandoned BLOBs, but this job only removes the blobs that have no content associated. I.e. the content is deleted by IContentRepository.Delete but the blob was left behind. The job uses the log to find out which content were deleted, then find those blobs.

How’s about the assets that still have contents associated with them, but not used anywhere? Time to get your hands dirty!

Due to the nature of this task, it is best to make it a scheduled job.

All of the assets are children under the global asset root. By iterating over them, we can check if each of them is being used by another content. If not, we will add them to a list for later delete. Before deleting the content, we will find the blob and then delete it as well. Easy, right?

To get the content recursively we use this little piece of code

        public virtual IEnumerable<T> GetAssetRecursive<T>(ContentReference parentLink, CultureInfo defaultCulture) where T : MediaData
        {
            foreach (var folder in LoadChildrenBatched<ContentFolder>(parentLink, defaultCulture))
            {
                foreach (var entry in GetAssetRecursive<T>(folder.ContentLink, defaultCulture))
                {
                    yield return entry;
                }
            }

            foreach (var entry in LoadChildrenBatched<T>(parentLink, defaultCulture))
            {
                yield return entry;
            }
        }

        private IEnumerable<T> LoadChildrenBatched<T>(ContentReference parentLink, CultureInfo defaultCulture) where T : IContent
        {
            var start = 0;

            while (!_isStopped)
            {
                var batch = _contentRepository.GetChildren<T>(parentLink, defaultCulture, start, 50);
                if (!batch.Any())
                {
                    yield break;
                }
                foreach (var content in batch)
                {
                    // Don't include linked products to avoid including them multiple times when traversing the catalog
                    if (!parentLink.CompareToIgnoreWorkID(content.ParentLink))
                    {
                        continue;
                    }

                    yield return content;
                }
                start += 50;
            }
        }

And we will start from SiteDefinition.Current.GlobalAssetsRoot, and use IContentRepository.GetReferencesToContent to see if it is used in any content (both CMS and Catalog). If not, we add it to a list. Later, we use IPermanentLinkMapper to see if it has any blob associated, and delete that as well

            foreach (var asset in GetAssetRecursive<MediaData>(SiteDefinition.Current.GlobalAssetsRoot, CultureInfo.InvariantCulture))
            {
                totalAsset++;
                if (!_contentRepository.GetReferencesToContent(asset.ContentLink, false).Any())
                {
                    toDelete.Add(asset.ContentLink.ToReferenceWithoutVersion());
                }

                if (toDelete.Count % 50 == 0)
                {
                    var maps = _permanentLinkMapper.Find(toDelete);
                    foreach (var map in maps)
                    {
                        deletedAsset++;
                        _contentRepository.Delete(map.ContentReference, true, EPiServer.Security.AccessLevel.NoAccess);
                        var container = Blob.GetContainerIdentifier(map.Guid);
                        //Probably redundency, can just delete directly
                        var blob = _blobFactory.GetBlob(container);
                        if (blob != null)
                        {
                            _blobFactory.Delete(container);
                        }
                        OnStatusChanged($"Deleting asset with id {map.ContentReference}");
                    }
                    toDelete.Clear();
                }
            }

We need another round of delete after the while loop to clean up the left over (or if we have less than 50 abandoned assets)

And we’re done!

Testing this job is simple – uploading a few assets to your cms and do not use it anywhere, then run the job. it should delete those assets.

Things to improve: we might want to make sure only assets that created more than a certain number of days ago are deleted. This allows editors to upload assets for later uses without having to use them immediately.

The code has been open sourced at vimvq1987/DeleteUnusedAssets: Delete unused assets from an Optimizely/Episerver site (github.com) , and I have uploaded a nuget package to Packages (optimizely.com) to be reviewed.

Skip validating carts

Carts are meant to be validated. Prices changed, customers add more quantity than allowed, promotions expired, stock ran out, etc.. All kinds of stuffs that make the items in carts need validation to make sure they up to date, and be ready to be converted to an order.

However, there are cases when you don’t want your carts to be validated. The most common case is of course, wish list – a special cart that allows customer to add items to, just to keep track of. You certainly don’t want to touch it. Another example is quote – when you give a specific item at specific price for a customer, and you don’t want it to be automatically changed to the public prices, which is different from that said price.

By default, when it is called to validate a cart, these things will be done:

Remove items that no longer available (either deleted or end of line)
Update prices of the items to the latest applicable prices, or remove items that have no prices.
Update quantity of the items (to comply with the settings or in stock quantity), or remove items that are out of stock
Apply promotions
Update quantity again (As promotions could do things like adding free items to the cart)

There are two ways you can avoid carts being validated, let’s see what we can do.

The “Wish list names” route

With OrderOptions you can set certain wishlist names to be exempted from the validation, using WishListCartNames. By default, it’s only “Wishlist”, but you can set several using the comma separator, like this

orderOptions.WishListCartNames = "Wishlist,Quote";

However there is a caveat, with this approach, carts in those names will not be shown in the Order Management (If you want, you can change that, however it is not an easy or quick one)

The OrderValidationService route

The validation of carts (or rather, order types in general) is done by OrderValidationService. And that class is meant to be extended if necessary. Here is how you would avoid validation carts with name “Quote”, using OrderValidationService

    public class CustomOrderValidationService : OrderValidationService
    {
        public CustomOrderValidationService(ILineItemValidator lineItemValidator, IPlacedPriceProcessor placedPriceProcessor, IPromotionEngine promotionEngine, IInventoryProcessor inventoryProcessor, OrderOptions orderOptions) : base(lineItemValidator, placedPriceProcessor, promotionEngine, inventoryProcessor, orderOptions)
        {
        }

        public override IDictionary<ILineItem, IList<ValidationIssue>> ValidateOrder(IOrderGroup orderGroup)
        {
            if (orderGroup.Name.Equals("Quote", System.StringComparison.OrdinalIgnoreCase))
            {
                return new Dictionary<ILineItem, IList<ValidationIssue>>();
            }
            return base.ValidateOrder(orderGroup);
        }
    }

And as OrderValidationService is registered by ServiceConfiguration attribute, you can register yours by

services.AddTransient<OrderValidationService, CustomOrderValidationService>();

One caveat though, making changes to OrderValidationService means those changes will apply to the entire website, so make sure the changes are actually the ones you want site-wide, not just in specific places.

Index only Catalog content

If you are using Find to index your content, you likely have used the Find Indexing job – which would index everything in one go. Today I stumped upon this question – A way to run indexing job for Commerce only | Optimizely Develope – and it is a good one – if you have many of content in CMS side, and they don’t change that often, if at all – you certain don’t want to waste time and resource in trying to reindex them again. Is there away to just index catalog content?

Yes, there is. It is a bit hacky solution, but it can certain work. But first, let’s dive in on how Find indexing job does it work. It relies on IIndexingJobService , which itself relies on ContentIndexer to do the job. In its turn, ContentIndexer uses a list of IReindexInformation to know which content to index, and in which languages. Here’s what it looks like

    public interface IReindexInformation
    {
        /// <summary>
        /// Content links to be reindexed.
        /// </summary>
        IEnumerable<ReindexTarget> ReindexTargets { get; }

        /// <summary>
        /// Gets the root to index.
        /// </summary>
        ContentReference Root { get; }
    }

It has one Root, and multiple ReindexTarget, which contains

    public class ReindexTarget
    {
        /// <summary>
        /// The content references.
        /// </summary>
        public IEnumerable<ContentReference> ContentLinks { get; set; }

        /// <summary>
        /// The languages the collection of <see cref="ContentReference"/> are enabled on.
        /// </summary>
        public IEnumerable<CultureInfo> Languages { get; set; }

        /// <summary>
        /// The site that the collection of <see cref="ContentReference"/> appears on
        /// or <c>null</c> if unknown.
        /// </summary>
        public SiteDefinition SiteDefinition { get; set; }
    }

As you might have guessed, Commerce has its own IReindexInformation to index catalog content. If we can only use that to run our job. This is how our “hack” begins

The interface IContentIndexer has no method to control the IReindexInformation`, but the default implementation ContentIndexer does. We set it to the only one we need, so here it is

        List<IReindexInformation> targets;
        var contentIndexer = _contentIndexer as ContentIndexer;
        if (contentIndexer != null)
        {
            targets = contentIndexer.ReindexInformation.ToList();
            var commerceReIndexInformation = targets.FirstOrDefault(x => x.GetType() == typeof(CommerceReIndexInformation));
            contentIndexer.ReindexInformation = new List<IReindexInformation>() { commerceReIndexInformation };
            _indexingJobService.Start(OnStatusChanged);

            contentIndexer.ReindexInformation = targets;
        }

A note is that you will still see the “Indexing Global assets and other data” message, because IIndexingJobService implementation will go through all SiteDefinition regardless and show that message, but the internal ContentIndexer will skip if the SiteDefinition passed to it does not match the SiteDefinition in the IReindexInformation (and for CommerceReIndexInformation it’s SiteDefinition.Empty

As I mentioned in the beginning, this is a bit hacky solution, as you have to cast IContentIndexer to its concrete implementation. The proper solution would be implement IContentIndexer yourself. Given that’s not a trivial job, I’ll leave at that.

Loading the contacts/organizations, the right way

If you have been using Business Foundation, you most likely know about a limitation – you can only load the first 1000 objects using the GetXXX methods. For example, by using CustomerContext.Current.GetOrganizations(), you can load the first 1000 organizations. In theory, you can get more objects by changing the value of MaxObjectsList. However, changing that has consequences. Changing that will affect all types of objects, including contacts, organizations, and your custom objects. Also, loading too much in one go is almost never a good idea.

Is there a better way?

Yes, of course – which is why we have this blog post

There is a “hidden” method from base class of Business Foundation – BusinessManager that takes paging parameters

public static EntityObject[] List(string metaClassName, FilterElement[] filters, SortingElement[] sorting, int? start, int? count)

You will need to convert the results to the type you want. Note that all Business Foundation objects are inherited from EntityObject. So if you want to get the contacts by paging, it would look like this:

                var contacts = BusinessManager.List(ContactEntity.ClassName, new FilterElement[0], new SortingElement[] { new SortingElement(sortField, sortType) }, startIndex, recordsToRetrieve)
                .OfType<CustomerContact>();

Let’s go through the parameters one by one.

The first you need is the class name of your objects. For contacts, you can use ContactEntity.ClassName as shown above. For organizations, OrganizationEntity.ClassName
Next one is the filter. As you are trying to load all objects, you can just pass in an empty (but not null) instance – new FilterElement[0]
Third one is how you want to sort it. If you pass an empty array, it will be sort by default. If you want to sort by Name for example, set your sortField to Name and sortType to one of SortingElementType (Asc or Desc)
Forth and fifth ones are what we are looking for, they’re simply paging parameters – which position to start getting, and how many objects to get. Combine this with a simple while loop, you can get all of your Business Foundation objects.

And that’s about it, my friends.

What’s about caching?

Caching with list is always tricky – as you have to keep track of each item in the list to make sure you invalidate the list cache if one of the item is changed (updated/removed). For the purpose of just loading all contacts/organizations, it is probably better to just skip caching, for simplicity.

Delete property no longer available in code

Recently I stumped upon this question Removing a property that no longer exists in the code (optimizely.com) . it’s a valid (and even good) question. It is easy to add a new property to your catalog content type – you can simply add a new property to the model, build and start the site. However the opposite is not easy. In Commerce 14 at least.

A property for the strongly typed content type, is actually mapped and backed by a MetaField in MetaDataPlus system (of course unless you specifically tell it not to, by using IgnoreMetaDataPlusSynchronization attribute). When you add a new property to your content type, build and start your site, your content type is scanned and metafields will be created if necessary. However, if you delete a property from your content type, the scanner will just leave the metafield there. There are a few reasons for that. Firstly, it allows loosely typed content type, i.e. content types with none, or only a few property defined. If you have used some kind of external PIM, you’ll understand why it is important. Lastly, because the property can be mapped with a metafield of different name, the scanner might have trouble figuring out which metafield to delete. All in all, keeping the metafields is the sensible (if not the right) choice.

Then what to do if you want to delete the property and also clean up the metafield? With Commerce 13 and earlier, you can detach a MetaField from its MetaClass(s), then delete it using Commerce Manager. With the dead of CM in Commerce 13, what is your option?

By using code, of course. There are a few APIs – namely MetaField and MetaClass that can be used for that purpose. Note that there are two MetaField and MetaClass, and only the ones in Mediachase.MetaDataPlus.Configurator namespace are what we want (the others are for Business Foundation)

Enough for chit chat, this is the code that you would need to run

        private void DeleteMetaField(string metafieldName)
        {
            var metaField = MetaField.Load(CatalogContext.MetaDataContext, metafieldName);
            if (metaField == null)
            {
                return;
            }
            foreach (int metaClassId in metaField.OwnerMetaClassIdList)
            {
                var metaClass = MetaClass.Load(CatalogContext.MetaDataContext, metaClassId);
                if (metaClass == null)
                {
                    metaClass.DeleteField(metafieldName);
                }
            }
            MetaField.Delete(CatalogContext.MetaDataContext, metaField.Id);
        }

It is pretty straightforward. We load the MetaField by its name, if it is not null, then we remove it from all MetaClass that are using it, then eventually delete it.

In beginning of this post we mentioned strongly typed content type, but note that order system also uses the same metaclass/metafield system, so this code can be used for them as well.

This piece of code can be used in an admin-privilege controller to delete metafields on demand. Until Commerce 14 allows you to do it with a proper UI.

Storing 100.000 prices per SKU – part 1

One of the questions I have received, from time to time, is that how to store a lot of prices per SKU in Optimizely (B2C) Commerce Cloud. While this is usually a perfect candidate for Optimizely B2B Commerce, there are many customers invested in B2C and want to make the best out of it. Is it possible?

It’s important to understand the pricing system of Optimizely Commerce (which is, written in detail in my book – shameless plug). But in short:

There are two price systems, IPriceService and IPriceDetailService
One is handling prices in batch – i.e. prices per SKU (IPriceService), and one is handling prices per individual price (IPriceDetailService)
Both are cached in latest version (cache for IPriceDetailService was added in late 13.x version)

With that in mind, it would be very problematic if you use IPriceService for such high number of prices per SKU, because each time you save a price, you save a lot of prices at once (same as loading prices). This is how the default IPriceService implementation saves prices of a SKU

create procedure dbo.ecf_Pricing_SetCatalogEntryPrices
    @CatalogKeys udttCatalogKey readonly,
    @PriceValues udttCatalogEntryPrice readonly
as
begin
    begin try
        declare @initialTranCount int = @@TRANCOUNT
        if @initialTranCount = 0 begin transaction

        delete pv
        from @CatalogKeys ck
        join dbo.PriceGroup pg on ck.CatalogEntryCode = pg.CatalogEntryCode
        join dbo.PriceValue pv on pg.PriceGroupId = pv.PriceGroupId

        merge into dbo.PriceGroup tgt
        using (select distinct CatalogEntryCode, MarketId, CurrencyCode, PriceTypeId, PriceCode from @PriceValues) src
        on (    tgt.CatalogEntryCode = src.CatalogEntryCode
            and tgt.MarketId = src.MarketId
            and tgt.CurrencyCode = src.CurrencyCode
            and tgt.PriceTypeId = src.PriceTypeId
            and tgt.PriceCode = src.PriceCode)
        when matched then update set Modified = GETUTCDATE()
        when not matched then insert (Created, Modified, CatalogEntryCode, MarketId, CurrencyCode, PriceTypeId, PriceCode)
            values (GETUTCDATE(), GETUTCDATE(), src.CatalogEntryCode, src.MarketId, src.CurrencyCode, src.PriceTypeId, src.PriceCode);

        insert into dbo.PriceValue (PriceGroupId, ValidFrom, ValidUntil, MinQuantity, MaxQuantity, UnitPrice)
        select pg.PriceGroupId, src.ValidFrom, src.ValidUntil, src.MinQuantity, src.MaxQuantity, src.UnitPrice
        from @PriceValues src
        left outer join PriceGroup pg
            on  src.CatalogEntryCode = pg.CatalogEntryCode
            and src.MarketId = pg.MarketId
            and src.CurrencyCode = pg.CurrencyCode
            and src.PriceTypeId = pg.PriceTypeId
            and src.PriceCode = pg.PriceCode

        delete tgt
        from dbo.PriceGroup tgt
        join @CatalogKeys ck on tgt.CatalogEntryCode = ck.CatalogEntryCode
        left join dbo.PriceValue pv on pv.PriceGroupId = tgt.PriceGroupId
        where pv.PriceGroupId is null

        if @initialTranCount = 0 commit transaction
    end try
    begin catch
        declare @msg nvarchar(4000), @severity int, @state int
        select @msg = ERROR_MESSAGE(), @severity = ERROR_SEVERITY(), @state = ERROR_STATE()
        if @initialTranCount = 0 rollback transaction
        raiserror(@msg, @severity, @state)
    end catch
end

If you have experience with SQL (which you probably should), you will see that it’s a deletion of rows in PriceValue that have CatalogEntryCode same as , then a merge, then a deletion of left over rows. To make matters worse, IPriceService system stores data on 3 tables: PriceValue, PriceGroup and PriceType. Imagine doing that with a few dozen of thousands rows.

Even if you change just one price, all prices of that specific SKU will be touched. It’d be fine if you have like ten prices, but if you have ten thousands prices, it’ll be a huge waste.

Not just that. To save one price, you would still need to load all prices of that specific SKU. That’s two layers of waste: the read operations at database layer, and then on application, a lot of price objects will need to be constructed, and then you need to recreate a datatable to send all the data back to the database to do the expensive operation above.

And wait, because the prices saved to IPriceService needs to be synchronized to IPriceDetailService (however, you can disable this). Prices that were changed (which is, all of them) need to be replicated to another table.

So in short, IPriceService was not designed to handle many prices per SKU. If you have less than a few hundred prices per SKU (on average), it’s fine. But if you have more than 1000 prices per SKU, it’s time to look at other options.

Where to store big collection data

No, I do not mean that big, big data (in size of terabytes or more). It’s big collection, like when you have a List<string> and it has more than a few hundreds of items. Where to store it?

Naturally, you would want to store that data as a property of a content. it’s convenient and it just works, so you definitely can. But the actual question is: should you?

It’s as simple as this

public virtual IList<String> MyBigProperty {get;set;}

But under the hood, it’s more than just … that. Let’s ignore UI for a moment (rendering such long list is a bad UX no matters how you look at it, but you can simply ignore rendering that property by appropriate attributes), and focus on the backend aspects of it.

List<T> properties are serialized as a long strings, and save at once to database. If you have a big property in your content, this will happen every time you load your content:

The data must be read from database, then transferred through the network
The data must be parsed to create an array (the underlying data structure of List<T>. The original string is tossed away.
Now you have a big array that you might not use every time. it’s just there taking your previous LOH (as with the original string)

Same thing happens when you actually save that property

The data must be serialized as string, the List<T> is now tossed away
The data then must be transferred through the network
The data then saved to database. Even though it is a very long string and you changed, maybe 10 characters, it’s completely rewritten. Due to its size, there might be multiple page writes needed.

As you can see, it can create a lot of waste, especially if you rarely use that property. To make the matter worse, due to the size of the property, it means they are taking up space in LOH (large objects heap).

And imagine if you have such properties in each and every of your content. The waste is multiplied, and your site is now at risk of some frequent Gen 2 Garbage collection. Nobody likes visiting a website that freezes (if not crashes) once every 30 minutes.

Then when to store such big collection data?

The obvious answer is … somewhere else. Without other inputs, it’s hard to give you some concrete suggestions, but how’s about a normalized custom table? You have the key as the content reference, and the other column is each value of the list. Just an idea. Then you only load the data when you absolutely need it. More work, yes, but it’s the better way to do it.

Just a reminder that whatever you do, just stay away from DDS – Dynamic Data Store. It’s the worst option of all. Just, don’t 🙂

Getting being delete variant codes

I got the question here Content Events and Service API | Optimizely Developer Community , and CHKN contacted Optimizely developer support service which is the good/right thing to do. However it could be beneficial for a wider audience, hence this blogpost.

If you are developing new Optimizely Commerce Cloud now, you should be using the latest version (CMS 12/ Commerce 14), or at least Commerce 13.31 if you have to stay with .NET 4.8. You then could use the IContentEvents to listen to any events that might be fired from Service API.

However, if you are using older versions, you might be limited to lower level, non content event through EventContext. It works, but with a catch: there is no EntryDeleting event in EventContext. At this point I’m not entirely sure why, probably just an overlook. But it’s not impossible to work around that issue.

As I suggested in the post, EntryUpdating is like a blanket event – every change to the entries goes through there. The sender is a CatalogEntryDto, which should contain information about the being deleted entries.

        private void Instance_EntryUpdating(object sender, Mediachase.Commerce.Catalog.Events.EntryEventArgs e)
        {
            var dto = sender as CatalogEntryDto;
            if (dto != null)
            {
                var rows = dto.CatalogEntry.Where(x => x.RowState == DataRowState.Deleted);
                var deletingCodes = rows.Select(x => (string)x["Code", DataRowVersion.Original]);
                //do stuffs with codes being deleted.
            }
        }

And then to listen to the event in one of your IInitializationModule

EventContext.Instance.EntryUpdating += Instance_EntryUpdating;

However, there is a caveat here: as EntryUpdating happens before the entry deletion, it is possible that the change did not go through (i.e. the changes were to be reverted). It is unlikely, but it’s a possibility. You might accept that, or you can:

Store the id and code in a “Deleting entry” dictionary
Listen to EntryDeleted event and match from DeletedEntryEventArgs parameter (which contains EntryId ) to get the deleted Code, and continue from there.

Pro Optimizely Commerce Cloud can now be preordered

If you want to start your week with a purchase, then I have a (hopefully) good news for you: My book Pro Optimizely Commerce Cloud can now be preordered (but with immediate access to content). https://leanpub.com/prooptimizelycommercecloud

It took me a week due to an technical issue with Leanpub, but in the end it can now be purchased.

I planned to finish this book in August. As always, I overestimated my ability to work during nights, and underestimated my procrastination power. I made some progress, but far from completing the book.

My solution? Well, I will let money speak. By letting you (my dear readers) preorder the book, I make a commitment to finish it. The book is now at 50% complete, and I aim to make it content complete before the end of this year.

If you own the previous version (even if you got it for free), there will be a coupon with 60% off coming your way, please check your mailbox!

The book is dedicated to my children, Emmy and Erik. While I can’t say they are very supportive of me writing the book (quite the contrary), they are my biggest motivation to make more money (cough, cough), and are biggest smiles on my face.

I would like to thank many of my colleagues, former and current, who helped and supported me writing the first book, which was the beginning of everything.

I would like to thank Optimizely community for inspiration to write this book.

And of course, I would like to thank you, for buying and reading it.

And I hope it’ll help you.

The hidden gotcha with IObjectInstanceCache

It’s not a secret that cache is one of the most, if not the most, important factors to their website performance. Yes cache is great, and if you are using Optimizely Content/Commerce Cloud, you should be using ISynchronizedObjectInstanceCache to cache your objects whenever possible.

But caching is not easy. Or rather, the cache invalidation is not easy.

To ensure that you have effective caching strategy, it’s important that you have cache dependencies, i.e. In principles, there are two types of dependencies:

Master keys. This is to control a entire cache “segment”. For example, you could have one master key for the prices. If you needs to invalidate the entire price cache, just remove the master key and you’re done.
Dependency keys. This is to tell cache system that your cache item depends on this or that object. If this or that object is invalidated, your cache item will be invalidated automatically. This is particularly useful if you do not control this or that object.

ISynchronizedObjectInstanceCache allows you to control the cache dependencies by CacheEvictionPolicy . There are a few ways to construct an instance of CacheEvictionPolicy, from if the cache expiration will be absolute (i.e. it will be invalidated after a fixed amount of time), or sliding (i.e. if it is accessed, its expiration will be renewed), to if your cache will be dependent on one or more master keys, and/or one or more dependency keys, like this

   
        /// <summary>
        /// Initializes a new instance of the <see cref="CacheEvictionPolicy"/> class.
        /// </summary>
        /// <param name="cacheKeys">The dependencies to other cached items, idetified by their keys.</param>
        public CacheEvictionPolicy(IEnumerable<string> cacheKeys)

        /// <summary>
        /// Initializes a new instance of the <see cref="CacheEvictionPolicy"/> class.
        /// </summary>
        /// <param name="cacheKeys">The dependencies to other cached items, idetified by their keys.</param>
        /// <param name="masterKeys">The master keys that we depend upon.</param>
        public CacheEvictionPolicy(IEnumerable<string> cacheKeys, IEnumerable<string> masterKeys)

The constructors that takes master keys and dependency keys look pretty the same, but there is an important difference/caveat here: if there is no dependency key already existing in cache, the cache item you are inserting will be invalidated (i.e. removed from cache) immediately. (For master keys, the framework will automatically add an empty object (if none existed) for you.)

That will be some unpleasant surprise – everything seems to be working fine, no error whatsoever. But, if you look closely, your code seems to be hitting database more than it should. But other than that, your website performance is silently suffering (sometimes, not “silently”)

This is an easy mistake to make – I did once myself (albeit long ago) in an important catalog content cache path. And I saw some very experienced developers made the same mistake as well (At this point you might wonder if the API itself is to blame)

Take away:

Make sure you are using the right constructor when you construct an instance of CacheEvictionPolicy . Are you sure that the cache keys you are going to depend on, actually exist?

In newer version of CMS, there would be a warning in log if the cache is invalidated immediately, however, it could be missed, unless you are actively looking for it.

Note that this behavior is the same with ISynchronizedObjectInstanceCache as it extends IObjectInstanceCache.