Customer Stories
Any company can talk about themselves to make it sound like they're the greatest thing going. But when you need help, you need an expert, not a salesman.
You need to know that they company you choose has the knowledge and experience that they claim to have. That is why we have posted a few examples of how TL Tech has helped real customers solve real problems.
After a failed upgrade, the customer's webmail system was inaccessible. The hosting company escalated the ticket and got the help of the vendor of the hosting control panel. Over a week later, the vendor managed to fix the webmail system, but in the process lost all of the email--over 4 million email messages belonging to the customer's own clients were now gone. Furthermore, the client was no longer able to send and receive new email. After finding themselves unable to help the customer further, the hosting company referred the customer to TL Tech.
TL Tech discovered that the email messages were not completely lost; but rather all 4 million messages had been smashed into one single email message. All of the original mailboxes had been destroyed, preventing new mail from arriving.
To fix the problem, TL Tech wrote a program that extracted the mail messages from the one large file, then determined from the contents of the message which mailbox it belonged to, and finally re-created the original message in the appropriate mailbox. Using this custom-built program, we were able to rebuild the entire email store.
The mail system had been down for over a week before the customer was referred to TL Tech; we had them back up and running within 24 hours.
After a hardware failure, the customer attempted to restore his data from a backup, but the control panel system said that the backup was corrupted. Similar attempts with older backup files also failed with the same error. Unable to help, the customer's normal technical support company suggested that the customer contact TL Tech instead.
Upon examination, TL Tech found that the backup files were not corrupted. Instead, we discovered a flaw in the hosting control panel's backup and restore system that caused it to fail to load its own backup files.
To solve the problem, we wrote a program that systematically modified the backup file to make it compatible with the restore process. The customer had his server restored and working that very same day. TL Tech then notified the vendor of their backup flaw, which was subsequently fixed in a later version.
The customer had a set of servers that he intended to upgrade to new hardware. However, since the hosting provider does not perform server migrations, they recommended that the customer contact TL Tech instead.
Upon contacting TL Tech, the customer learned that we were also willing to show him how to perform the migrations himself--an option he much preferred. We walked the customer through the migration process, step by step, with the customer performing the migration on one of his servers while we guided him through the process. We also took additional time to answer all of his questions and to explain to him what precautions to take to ensure that throughout the whole process, no data is lost and his clients never see any interruption in service.
The customer then went on to perform the remaining migrations on his own, but called us back when he ran into a problem: one of the applications installed on an old server was improperly configured, such that if ever moved to another server it would not work correctly. We were able to quickly determine the cause of the problem, fix it, and explain to the customer what to look for to make sure he didn't have problems with that sort of thing in the future.
The customer needed a way of interacting with his server using a telephone. He had found some software that would allow his server to take telephone calls, but he couldn't make the software do what he wanted. He asked his hosting company for help, and they referred him to TL Tech.
TL Tech built for him a complete, customized telephone system for his server using Open Source tools and software. With our software, the server could take calls, respond to user requests, and even record messages. The project does exactly what he wanted, and costed him less than he would have paid for just one year's worth of licensing fees had he continued using the commercial product that he found first.
The customer needed to stream vast amounts of data to thousands of customers worldwide, with less than a tenth of a second delay end-to-end. Their existing solution regularly crashed, occasionally got up to 15 minutes behind, and needed to be rebooted daily to prevent data corruption. Based on past experience, they knew that they could get the help they needed from TL Tech.
We designed, built, tested, and then deployed a complete custom solution for the customer. The resulting solution scaled cleanly with their hardware. That is, adding twice as many servers would allow the company to support roughly twice as much data or twice as many clients. The total processing delay to decode, categorize, compress, and then distribute the data was less than 5 milliseconds. But the customer's favorite part about the solution was that it required absolutely no maintenance: it never crashed, and ran for years without ever rebooting.
The customer found that as his business grew, his database-centric application grew increasingly slow and unusable. He feared that someone malicious was interfering with his web application. His technical support company found nothing wrong, and suggested that he call TL Tech for more help.
We found no signs of intrusion or malware on the customer's server. However, after an examination of the customer's custom-built application and database, TL Tech found a few small changes that could drastically improve performance. After making these changes, we noted that requests that before took up to 3 minutes to process would now finish in less than a second. Web pages that had become completely unavailable now worked exactly as the company had hoped.
The customer had a custom-built application that used a set of pre-built libraries from a third-party vendor. Under rare but predictable circumstances, the whole server would crash. The third-party vendor insisted that their code was reliable, but the company could not find any flaws with their own code. As an existing TL Tech customer, the company immediately thought to come to us for help.
Since the third-party vendor would not provide the source code to to their library, we had to do all of our debugging work in assembly code. We determined the location of the fault, and then using our knowledge of compilers and assembly code, we were able to determine what code had produced the fault. We determined that it was a punctuation mistake in a specific looping operation in the third-party vendor's library. We were even able to determine from the purpose of the code, what the name of the faulty function would probably be, what the faulty code would look like, and approximately how far down the page to find it.
The customer delivered this information to the third-party vendor who, when presented with such specific details as to the nature and location of the error, were able to release a fixed version that same day.
The customer was using hard drive mirroring technology between two servers for redundancy, but after one of the servers failed, the system incorrectly began to destroy the data on the "good" hard drive. The customer noticed the failure and stopped the process, but the drive had already been corrupted and was unreadable. The customer was referred to TL Tech for help.
The "logical volume" and "partition" information on the drive had been overwritten, essentially destroying the index of what parts of the drive were being used for what. However, since we had a mirrored copy of the drive (albeit an old one), we were able to use information obtained from the copied drive to serve as landmarks on the drive we wanted to recover.
After finding the appropriate landmarks on the hard drive, we then ran the calculations backwards to determine where our data was. We then copied the important chunks of the hard drive to another server for analysis, and by the end of the day, the customer had all of his client's data.