Microsoft

Hung jury disagrees as to whether NT pro got caught sleeping

IT pros get busy. It's part of the job. What happens when things go wrong, though? Is it the IT pro's fault, or just another day in the life? Learn what TechRepublic members had to say about one situation.


Back in the spring, TechRepublic published an article presenting one IT professional’s experience. The writer was converting a broker’s management network from UNIX to Windows NT, but things ran awry.

We asked what you thought. Many TechRepublic members responded. Here’s what you had to say.

The situation
Our IT professional was preparing a Windows NT conversion from a UNIX system in an environment supporting 50 seats. The server was tested (minus critical applications) and it performed fine.

During implementation, a crisis arose when a tape backup didn’t work properly. In fact, the tape got eaten.

No fresh NT system backup was prepared or ready. You see, the backup was from a UNIX system, and a conversion needed to be run. Three days lapsed before the system became operational again.

Rookie mistake?
Should an extra tape have been made? Did our NT professional make a rookie mistake? Perhaps he was overwhelmed with the services that he had promised to provide within a 12-hour period. Here are some of the edited comments received from TechRepublic members.

From: Wade Stankich: ”I have been lucky so far to have a valid backup during mission-critical installations. But, I do know the frustration of telling our user base 'three hours max!' and then working through the night. Things do go wrong, and I do not see this as being a rookie move, just a mistake. We all make them from time to time.”

TechRepublic member Imran Syed said, ”I think the rookie mistake in this case was conducting a major conversion during working hours and giving a 12-hour time frame. First of all, a 12-hour time frame for a major conversion is not practical. First rule of networking: any upgrade or conversion should be done on weekends or off-hours to minimize the downtime. Give yourself enough space for mistakes, because things do go wrong.”

Here’s what Mark Douglas had to say. ”It would sound from the article that the individual involved was under extreme pressure from his bosses and clients. Nothing new there. He decided to skip the backup stage of the new installation due to what he had identified as another step that would delay the implementation of the new system. What he should have done was get the client to sign off on the skipping of this step after fully explaining the risks, and then gone ahead and done the backup anyway.”

Bill Williams said this was ”not a rookie mistake. He made a decision to skip a step in hopes of saving time. Sometimes that works, sometimes not.”

Rules of migration and financial persuasion
Should our NT pro have taken a few more precautionary steps before taking down the old system and putting up the new system? One member believes so, and gives a list of rules he always follows:

Jim Harring said IT people should realize they are not only performing a migration but managing expectations of how that migration will occur.

Here are his rules of migration:
  1. Whenever dealing with mission-critical data, have no less than two copies of the identical data. The old tape drive usually has lots of hours on the heads, so the likelihood of various errors is higher.
  2. Keep the old system available (but not online) until well after the new system is up.
  3. Build that time taken by step number one into your estimation of downtime.
  4. When management balks at the amount of downtime, practice "financial persuasion.”

Murphy’s Law rears its head
Perhaps our NT pro had all the odds stacked against him. One member suggests that Murphy’s Law may have had something to do with the problem:

Lowell Shim suggested that he doesn’t ”think the administrator made rookie mistakes. All that happened was that guy Murphy. No matter how good or experienced an administrator is, Murphy will prevail. All it takes is just one small slip and the situation ends up being a job from hell. Always after the fact you will wish you had done this or that differently.”

They sympathize completely
It seems that every great IT pro has had a bad day. Here are a few examples that TechRepublic members gave when they faced similar experiences:

Brian Beck said, ”I have made mistakes like this before. In today’s business, people want things done quickly. However, quickly usually costs more money. I have had to battle with my boss and other personnel about how long it may take to get a job done. They finally figured out that a lot of things in the IT business just couldn’t be hurried. Once they understood that point, it made my job a lot less stressful.”

According to TechRepublic member Bob Cloninger, ”Very few of us have nerves of steel when the users are getting restless and the job has already gone longer than expected. How many times have I sworn not to sacrifice technical necessity to political expediency?”

Joescat may have summed it up best: ”Yep, been there and done that, having not taken the time to draw safety lines to guide the way back. But, who really expects something like an eaten tape, just when you need to restore?”
Have you had an upgrade go terribly wrong? If so, let us know! Feel free to leave a post below, or send us a note.
Back in the spring, TechRepublic published an article presenting one IT professional’s experience. The writer was converting a broker’s management network from UNIX to Windows NT, but things ran awry.

We asked what you thought. Many TechRepublic members responded. Here’s what you had to say.

The situation
Our IT professional was preparing a Windows NT conversion from a UNIX system in an environment supporting 50 seats. The server was tested (minus critical applications) and it performed fine.

During implementation, a crisis arose when a tape backup didn’t work properly. In fact, the tape got eaten.

No fresh NT system backup was prepared or ready. You see, the backup was from a UNIX system, and a conversion needed to be run. Three days lapsed before the system became operational again.

Rookie mistake?
Should an extra tape have been made? Did our NT professional make a rookie mistake? Perhaps he was overwhelmed with the services that he had promised to provide within a 12-hour period. Here are some of the edited comments received from TechRepublic members.

From: Wade Stankich: ”I have been lucky so far to have a valid backup during mission-critical installations. But, I do know the frustration of telling our user base 'three hours max!' and then working through the night. Things do go wrong, and I do not see this as being a rookie move, just a mistake. We all make them from time to time.”

TechRepublic member Imran Syed said, ”I think the rookie mistake in this case was conducting a major conversion during working hours and giving a 12-hour time frame. First of all, a 12-hour time frame for a major conversion is not practical. First rule of networking: any upgrade or conversion should be done on weekends or off-hours to minimize the downtime. Give yourself enough space for mistakes, because things do go wrong.”

Here’s what Mark Douglas had to say. ”It would sound from the article that the individual involved was under extreme pressure from his bosses and clients. Nothing new there. He decided to skip the backup stage of the new installation due to what he had identified as another step that would delay the implementation of the new system. What he should have done was get the client to sign off on the skipping of this step after fully explaining the risks, and then gone ahead and done the backup anyway.”

Bill Williams said this was ”not a rookie mistake. He made a decision to skip a step in hopes of saving time. Sometimes that works, sometimes not.”

Rules of migration and financial persuasion
Should our NT pro have taken a few more precautionary steps before taking down the old system and putting up the new system? One member believes so, and gives a list of rules he always follows:

Jim Harring said IT people should realize they are not only performing a migration but managing expectations of how that migration will occur.

Here are his rules of migration:
  1. Whenever dealing with mission-critical data, have no less than two copies of the identical data. The old tape drive usually has lots of hours on the heads, so the likelihood of various errors is higher.
  2. Keep the old system available (but not online) until well after the new system is up.
  3. Build that time taken by step number one into your estimation of downtime.
  4. When management balks at the amount of downtime, practice "financial persuasion.”

Murphy’s Law rears its head
Perhaps our NT pro had all the odds stacked against him. One member suggests that Murphy’s Law may have had something to do with the problem:

Lowell Shim suggested that he doesn’t ”think the administrator made rookie mistakes. All that happened was that guy Murphy. No matter how good or experienced an administrator is, Murphy will prevail. All it takes is just one small slip and the situation ends up being a job from hell. Always after the fact you will wish you had done this or that differently.”

They sympathize completely
It seems that every great IT pro has had a bad day. Here are a few examples that TechRepublic members gave when they faced similar experiences:

Brian Beck said, ”I have made mistakes like this before. In today’s business, people want things done quickly. However, quickly usually costs more money. I have had to battle with my boss and other personnel about how long it may take to get a job done. They finally figured out that a lot of things in the IT business just couldn’t be hurried. Once they understood that point, it made my job a lot less stressful.”

According to TechRepublic member Bob Cloninger, ”Very few of us have nerves of steel when the users are getting restless and the job has already gone longer than expected. How many times have I sworn not to sacrifice technical necessity to political expediency?”

Joescat may have summed it up best: ”Yep, been there and done that, having not taken the time to draw safety lines to guide the way back. But, who really expects something like an eaten tape, just when you need to restore?”
Have you had an upgrade go terribly wrong? If so, let us know! Feel free to leave a post below, or send us a note.

Editor's Picks

Free Newsletters, In your Inbox