Migration duration

Migration duration varies between projects and depends on multiple technical factors. It cannot be accurately estimated from account count or total storage alone.

The main reason for that is that service providers that host the accounts enforce API quotas that limit how many items can be read from the source and created in the destination within a given time window. These limits vary depending on item counts, total account size, content types, and account activity, and they can change dynamically due to throttling or provider-side performance conditions.

This article explains how VaultMe for Admins generates time estimates, what factors most strongly affect the migration speed, and how to plan a migration timeline based on the expected behavior of user accounts.

It also covers how parallel migrations work, how to reduce the likelihood of delays, and what to do if an account slows down or appears to stall.

How do I get a timeline estimation for my project?

The most reliable way to get a timeline estimate is to set up a draft migration in VaultMe for Admins.

Once all accounts are connected and scanned, VaultMe for Admins can analyze the scope of the migration and produce an estimation based on patterns observed in many previous migrations with similar characteristics.

High-level details alone, such as the number of accounts or total storage usage, do not provide enough context to produce a meaningful estimate because they do not reflect account composition (the number of items and their individual and total sizes) or account activity.

The sections below explain what factors affect the migration speed and how to plan a migration timeline.

What affects the migration speed?

VaultMe for Admins performs migrations using standard, publicly available APIs offered by the service providers that host the accounts. These APIs enforce strict multi-factor thresholds on many operations, which directly limit the maximum possible migration speed. VaultMe for Admins is designed to migrate extremely large amounts of content extremely quickly; however, its performance is limited by these thresholds.

Both reading from the source and writing to the destination can become bottlenecks, as both the source and destination providers can limit the performance on accounts.

In practice, writing to the destination tends to take longer than reading, so the performance of the destination accounts is often the main factor that determines overall migration speed. As a result, even when items can be read quickly from the source, the migration often has to wait for the destination to process and accept them.

Key factors affecting migration speed are explained below.

Per-minute limits on the service providers' end

All APIs enforce limits on the maximum number of items that can be read or created within a single minute. This means that the number of items that can be added to the destination account per minute is always limited.

In practice, this limitation is the key factor that defines the maximum available migration speed. For accounts with small item counts, this is often the only limit that meaningfully affects migration speed. In this context, the number of items, not their size, is the defining factor.

In some situations, service providers may temporarily tighten this limit, most commonly due to increased recent or parallel activity. This does not stop the migration but slows down the process, so fewer items can be added to the destination account over a set period. This type of load management is known as throttling.

The rules governing this type of throttling are not publicly documented and may involve AI-based decision mechanisms on the service provider's end.

Daily limits on the service providers' end

In addition to per-minute limits, most service providers also enforce limits on the maximum number of items that can be created or downloaded over a longer period, typically 1 to 24 hours.

These limits may apply to all items or only to specific item types (such as calendars), and may also take into account the total size of all items. These limitations are primarily designed to protect accounts from abusive or automated behavior, but they affect any application that generates high load over time, including migrations.

When these limits are reached, migration is paused until the limits are reset.

Unlike per-minute limits, daily limits are non-linear. They are only minimally documented and are often triggered in ways that seem unexpected or unpredictable.

Daily limits play a critical role for large accounts (i.e., over 200,000 items or 200 GB of content) and especially for extra-large accounts (i.e., over 1 million items or 1 TB of content). Migrations of extra-large accounts typically take at least several days to a week or more to complete.

While a timeline of a week or longer may seem discouraging, migrations performed by VaultMe for Admins do not cause downtime. Accounts remain active throughout the migration, and users can continue working with their content. All changes made by users during the process are automatically synced at the end of the migration.

Account size and composition

Because API limits are not linear and take into account multiple dimensions at the same time, the composition of the accounts has a direct effect on the migration time.

The number of items

Even if accounts appear small in terms of storage use (e.g., a 20 GB mailbox), a high item count (e.g., 500,000 emails) can cause the migration to hit the API limit or experience temporary throttling, sometimes multiple times during the migration.

When this happens, VaultMe for Admins has to pause the migration and wait for the limitations on the service providers' end to be lifted.

For accounts with large item counts, the number of items affects migration time more than the account size.

The total size of the content

Total size is more important when individual items are very large. For example, extra-large drives may contain fewer items, so the item limit is rarely reached. However, large files often contribute to reaching other thresholds, such as the maximum amount of content that can be added to an account over a period of time.

Combination of item count and content size

Quotas applied by service providers often take into account several dimensions at the same time. How various factors are combined is not fully documented. As a result, two accounts with the same total size or the same number of items can behave very differently during migration.

Typically, the most challenging accounts are often those with very high item counts, even if their total size is relatively small. Large mailboxes with hundreds of thousands or millions of messages are usually slower and less predictable than fewer-item, higher-volume file accounts.

Number of accounts in a batch

From VaultMe for Admins's perspective, the batch size alone does not reduce the migration speed or artificially slow individual accounts. However, it does have an indirect impact on how migrations behave in practice.

With a higher number of accounts, it is likely that one or more accounts will experience delays due to API limits, throttling, or other performance issues. When this happens, those accounts may pause or slow down independently, which in turn extends the total time needed for the entire project to complete.

Additionally, the more accounts are being migrated, the higher the load on the source and destination environments, which may also be a contributing factor to throttling issues.

Previous account activity

When allocating processing power to each account, service providers consider not only the load generated by VaultMe for Admins, but also the entire history of activity over the past few days.

If the accounts have been used heavily prior to a VaultMe for Admins migration, including extracting, migrating, or reorganizing content manually or automatically, their performance may already be limited by the time the migration begins.

VaultMe for Admins cannot predict or account for these limitations when generating an initial time estimate, but the estimate will be updated as more information about the accounts' performance becomes available.

Concurrent account activity

When allocating API resources, service providers estimate the total load observed in an account and enforce limits based on that combined activity.

This total load includes all user actions and third-party application activity. For this reason, any activity occurring in the accounts while the migration is running may reduce the migration speed because users and/or applications are competing for API resources.

Examples of concurrent activity include users actively working with their content, background synchronization, backup or archiving, eDiscovery processes, email security scanning, or other migration or automation processes running in parallel. All of these consume the same API resources.

For large migrations, minimizing concurrent activity during the migration helps reduce the likelihood of throttling and delays.

Broken or corrupted items

In some cases, migrations may encounter issues caused by corrupted or inconsistent items in the source accounts. Examples include items with invalid metadata, malformed attachments, or older items that no longer fully comply with the provider’s current technical requirements.

Destination service providers may reject requests to recreate these, and this may cause errors and retries that extend migration time. While VaultMe for Admins is designed to handle errors and continue wherever possible, repeated failures on problematic items can slow down the migration and contribute to throttling.

These issues are rare, but they are more likely to occur in older accounts or those that have undergone multiple migrations. Their impact on migration time is difficult to predict in advance.

Provider-side technical issues

Service providers occasionally experience unexpected technical issues that can affect account and migration performance. These issues are rare but can still affect the migration timeline.

Examples include temporary slowdowns in specific services, delays in processing the content, or intermittent API errors for certain item types.

In such cases, accounts may respond more slowly or inconsistently than they normally would, and VaultMe for Admins may encounter random errors. Error messages generated by service providers in cases like this are often generic and require manual review, which also extends the migration time.

In case of provider-side technical issues, VaultMe for Admins will retry operations or temporarily slow down the migration until the issues are resolved. These situations are outside VaultMe’s control and typically go away on their own without intervention.

Will all accounts be migrated at the same time?

VaultMe for Admins migrates accounts in parallel using multi-threading. There is no limit on the number of processing threads that would force accounts to wait in line before starting, so migrations for all accounts can be active at the same time.

The only limitation concerns pairs of accounts that use the same destination, such as when several mailboxes need to be migrated into one archival mailbox in the destination. In this scenario, one destination account can be used for only one migration at a time. The remaining migrations using the same destination will be automatically queued and processed sequentially.

At the same time, no two accounts are the same. Each account differs in total size, number of items, item sizes, folder structure, and performance. Because of this, migrations of individual accounts will not complete at the same time, even when they are started together.

How accurate are VaultMe's time estimates?

VaultMe for Admins's time estimates are projections rather than guarantees. These projections are based on statistics collected from many previous migrations with similar characteristics, such as the combination of service providers, content types, item sizes, and item counts. Because each account is unique, each migration is different.

Since API quotas are non-linear and depend on the specific composition of each account (including the mix of item sizes and item counts), migration time, even for similar accounts, can vary. Factors such as throttling, account performance, and API limits directly affect migration time and cannot be fully predicted in advance.

The estimates are reliable for setting expectations and comparing relative scope (for example, identifying which accounts are likely to take longer than others), but they should not be treated as precise timelines.

For planning purposes, it is best to allow buffer time, especially for large batches or accounts with very high item counts.

Initial time estimates may change, sometimes even multiple times during the migration, and this is totally normal. VaultMe for Admins will update the remaining time estimate as more information becomes available during the migration.

Can VaultMe speed up the process?

It is not possible to speed up the process because the limitations imposed by service providers cannot be bypassed by either VaultMe for Admins or any other third-party migration app.

VaultMe for Admins always attempts to complete the migration as quickly as allowed by service providers, and it continuously adjusts the initial estimate when more information regarding the performance of the accounts becomes available.

However, it cannot "push through" the API limits or reduce throttling without resorting to temporary pauses.

Pausing and unpausing the migration manually will not improve the migration speed and may sometimes even reduce it. The best course of action is to reduce account activity and to give the migration more time.

How do I plan the timeline?

When planning a migration timeline, it is important to account for the size and composition of accounts involved and the limitations that service providers may impose. Different account profiles behave differently during migration, and the slowest accounts typically determine the overall project duration.

Accounts with extremely high item counts, such as large mailboxes containing hundreds of thousands or millions of messages, usually present the highest time-related risk. Even when the total storage size is relatively small, these migrations tend to be slow because service providers typically allow adding only a limited number of new items within a certain period. As a result, migrations for such mailboxes often take one to two weeks to complete.

Extra-large drives measured in multiple terabytes may also require more time, but their behavior is often more predictable when item counts are low. Typically, several hundred gigabytes of content can be added to the destination over 24 hours, but these numbers vary from provider to provider.

The best practice is to identify the largest and most complex accounts in advance and treat them as pacing items for the project. Smaller accounts may complete much sooner, but overall completion should be planned around the accounts with the highest item counts or the largest size.

How do I reduce the possibility of delays?

Taking the following steps can reduce the possibility of delays:

1. Run the migration in consecutive smaller batches, if possible. Even though VaultMe for Admins processes accounts in parallel, smaller batches reduce the overall load on the source and destination accounts and lower the likelihood of triggering throttling across many accounts at the same time.
2. Start with a test group of accounts. Migrating a small, representative subset of accounts first helps check assumptions about item counts, account composition, and performance. This allows identifying potential bottlenecks early.
3. Minimize activity in the source and destination accounts during the migration. Reducing user activity and automated processes makes API capacity available for the migration itself. Examples of automated tools that compete for API capacity include email archiving tools, backup and eDiscovery solutions, email analytics or productivity tools, synchronization apps, and others. These applications often make heavy use of the same APIs and consume quotas, increasing the likelihood of throttling or API limits being reached during the migration.
4. Ensure that all destination accounts are ready before starting. This includes enabling all necessary services, assigning appropriate licenses, and setting appropriate storage quotas for all users. Preventable configuration issues are one of the most common causes of pauses during otherwise well-planned migrations.

Can users continue working in the accounts during the migration?

Users can continue working in their accounts while the migration is running. Normal user activity will not cause the migration to fail, and changes made in the accounts during the migration will be additionally copied at the end of the migration process.

However, active usage can affect migration speed. User activity consumes the same API resources as the migration, which may increase load and lead to slower progress or temporary throttling. For this reason, while continued work is supported, minimizing heavy activity during the migration is advised.

What should I do if my migration is stalled?

If a migration has not made any progress for a certain period, the migration page will show the "Delayed" status and a link to a detailed explanation of the account's performance.

The best solution is to leave the migration going and give it more time to complete. Temporary slowdowns and pauses are a normal part of large migrations and are usually caused by API limits and throttling rather than by errors.

If the performance of the accounts is limited by the service providers, VaultMe for Admins will continuously check the status and automatically restore normal migration speed once the performance is back to normal.

If a migration appears to be stalled for an extended period, it is usually best to review whether there is heavy concurrent activity in the source or destination accounts, or whether very large or high–item-count accounts are involved. In such cases, additional waiting time is expected.

Pausing and unpausing the migration manually does not improve performance and can sometimes extend the migration time.