How to Improve the Backup Success Rate of Data Centers?

Improve Backup Success Rate

According to industry analysts, a significant number of backup jobs (from 5 to 25%) are failing across various tiers of data centers. This impacts datacenters heavily on revenue loss, SLA-based penalties, and customer experience. Furthermore, loss of important data alienates customers and leads to low NPS (net promoter score).

Below image shows the current datacenter landscape with success rate and failure rate across various tiers. [Source: Gartner]

It is clear that the backup failure rate is least for tier 1 data centers at 5% and highest for tier 2 remote offices at 25%.

To reduce backup failure, it is necessary to find the root cause of the failure.

Following are a few frequent causes identified after monitoring some of the top datacenters.

Once the reason for failure is identified, the next step is to create a solution strategy to eliminate the problem and create successful backups. Below are some of the proven strategies to minimize backup failure.

Backup Audit tool

An audit tool helps data centers to conduct automated periodic audits of target systems and backup servers. Some of the features of audit tools are:

Firewall Configuration Audit: The audit tool runs a script at regular intervals to test the accessibility of the selected range of IP subnets, ports, target systems and backup servers. If any issue is found in the bidirectional communication, a ticket is raised automatically and sent to the concerned team.

Server Decommissioning: For servers that are consistently “inaccessible” and lead to backup failure, the audit tool automatically checks with the relevant internal application (Contract Management) for contractual information. If the application responds with a “contract expiry” message, the tool will initiate a ticket for decommissioning the server, disable further backups on it, and release the attached backup resources.

Backup Agent Audit: The audit tool runs a script on target systems and backup servers periodically to check if there is any mismatch in agent versions. If it finds a mismatch, a ticket gets generated automatically and sent to the admin team for customer approvals.

Database Permission Audit: Before running backup jobs, audit tool checks for access permission levels for the target system. If the tool identifies any gap, a ticket gets triggered to the admin team to get required permission from the customer for the associated target system.

Proactive capacity management

In order to effectively manage capacity and solve storage-related issues, data centers should focus on important areas to track and analyze various parameters such as location, libraries, media used, and storage-related information. Following measures would help data centers to reduce backup failures due to storage mismanagement.

Archival strategy for backup data

Primary storage is expensive as storage arrays are required to produce a sufficient level of input/output operations per second (IOPS) to meet operational requirements for user read/write activity. The data archives serve as an effective way for reducing primary storage consumption and related costs.

Cloud storage is a possible backup solution as it is cheaper and offers flexibility with an ongoing investment. Some solutions include Amazon Glacier, Microsoft StorSimple, and Google Drive.

Recommended archival strategy

Data classification: Classify the data with the most legal and regulatory exposure risk and update archival policy regularly for better compliance.

Well-defined retention policy: Align archival policy with different departments and BUs to retain information for varying time periods.

Tools for structuring the data: Use automation tools to structure data — indexing, auto-classification, text and content analytics helps to extract more value from data and store it efficiently.

Data centers can reduce approximately 25% of their storage costs by implementing an effective archival strategy. For Instance, focusing on applications that accumulate unstructured data such as audiovisuals or images and archiving them into the cloud will reduce the load on the primary storage.

Parallel processing of backup jobs

The following diagrams illustrate the difference between serial backup processing and parallel backup processing. In parallel processing, the same storage is logically divided into different instances. This makes parallel processing of jobs more effective and faster than the serial backup processing.

Conclusion

This article provides insights on how having a customized backup tool can minimize failure, improve storage performance, and achieve 99% successful backup.

By Vishwa Nigam, Manager – Business Analysis & Insights, Prodapt

Vishwa is an experienced business manager with a demonstrated history of building & delivering actionable insights on devops, IoT, and robotic process automation. His areas of interest are analytics, process improvement, and business model innovation.

Mark Greenlaw
Free Cloud Migrations are Expensive The cloud is becoming the primary place where work gets done. By 2025, Gartner estimates that enterprise spending on public cloud computing will overtake traditional IT hardware. Why? One reason ...
Gilad David Maayan
What Is Cloud Deployment? Cloud deployment is the process of deploying and managing applications, services, and infrastructure in a cloud computing environment. Cloud deployment provides scalability, reliability and accessibility over the internet, and it allows ...
Matt Hallett
Data Clean Rooms are Changing the Game for Marketers It’s no surprise that data clean rooms (DCRs) have become the go-to solution for customer insights. With the depreciation of cookies and growing concerns about data ...
Louis
More CISOs will have to deliver revenue growth to protect their budgets and grow their careers in 2023 and beyond, and a core part of that will be getting multicloud security right. It’s the most common infrastructure strategy for ...
Josh Miramant
Benefits of a Data-First Culture When it comes to analytics solutions, centralization versus decentralization is one constant tension that’s plagued data architects for years now. Both options offer their own sets of advantages and disadvantages, ...
Stacey Farrar
Modern Auth and Exchange Online Migrations Microsoft has phased out Basic Authentication (Basic Auth), replacing it with Modern Authentication (Modern Auth) to provide increased protection and user security. Through this, Microsoft has turned off Basic ...
Sofia Jaramillo
Augmented Reality in Architecture Augmented reality (AR) is a growing field of study and application in the world of architecture. This useful tool can help us visualize architectural designs by superimposing them onto real-world scenes ...
Patrick Melampy
Cloud On-Ramp and Protecting Performance The expansion of remote work and the massive growth in usage of cloud-based applications have stressed existing infrastructure and put a keen focus on the performance of everyone’s network environment ...