100% CPU Load

I was about to help my son-in-law move a large chair at his and my daughter's home when the BlueTooth earbud buzzed in my ear. I pressed the button to take the call. It was a familiar client from Alabama.

"We need some help. Our CPU is pegged at 100% busy."

My son-in-law stepped away from the chair so that I could take the call. I waved him back, but he ignored me. I asked the client to run a couple of "ps -e" commands and tell me whether the number of processes was the same. It was. I then asked him to run two "ps -e"'s, waiting about 10 seconds between, and redirecting them to different files. While he was doing that I signaled to my son-in-law and said "I can lift this while I talk".

"Huh?", said the voice in my ear. I explained that I was talking to someone else and asked him to run a "diff" on the two files he had created. We picked up the chair as he did that.

"It's the getty lines that are different. The getty's on the Digiboard are accumulating CPU time", my client said.

I was now walking backward down a hall. "Getty's shouldn't be accumulating time", I said. "Is anyone trying to login? Is it all the Digi's?"

My son-in-law surely understood my end of the conversation, but my daughter was giving me those "Dad is a geek" looks. This is how I earn my living, honey. It's what put you through college."

The client answered. "Nope. Nobody is trying to log in. It's all on one bank, the 't' bank. Accumulating lots of time."

"OK", I said as I backed into the room that was to be this chair's new home, "let's disable all the 't' ports". I thought about how many different rooms this chair had been in in its history. It was hand built by one of my ancestors sometime in the late 1700's or early 1800's, so it has been around. It's been rebuilt a few times, so there may not be too much left of the original, but it's still old and deserves respect. We put the chair down in a corner. It looked happy there, if a chair can look happy.

"All disabled", my client reported.

"Let's see another 'ps'", I requested.

I was signaling to my daughter that we were all done and I was leaving. We waved goodbye and I left the house.

After a moment the client said, "The getty's are still accumulating time".

I was walking to my car now. I knew what the problem was. I warned the client to call back if we got disconnected as I drove; there are some cellular troubles spots on the road home. I told him what the problem was (though I think he had realized it himself at this point) and told him what to do. What would you have said at this point?

Scroll down for the answer.
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Obviously there was a serious hardware problem in the 't' bank of the multiport card. After disabling, the getty's shouldn't have still been running because "init" kills processes that it started - either "disable" was broken, or the processes were stuck in the driver so "init" was unable to kill them. I had the client do a "grep 'tty[tT]' /etc/inittab" and confirm that no lines still said "respawn". As none did, the processes were stuck waiting for some hardware event to wake them up (see Why can't I kill a process with -9?). As it was only the 't' bank, presumably neither the driver nor the kernel were at fault here, so the most likely cause was simply bad hardware.

Of course it could be some odd memory problem or really strange driver bug, but that's very unlikely. He rebooted, and the CPU load went back to normal levels. He said he could take it from there (replacing the hardware or looking for faulty cables, power supplies, etc.)

The http://aplawrence.com/Unixart/slow.html article at this site discusses various tips for diagnosing this type of problem.

Sun Feb 12 05:51:26 2006: Subject:   FloralMike


And my wife says that men cannot multi-task. Ha!




Sun Feb 12 11:57:29 2006: Subject:   TonyLawrence


Your wife is probably right. I might have walked almost anywhere if my daughter hadn't been waving me in the right direction.



Mon Feb 13 20:54:00 2006: Subject:   Niall


I wasn't ignoring you, I was obeying your daughter. Always a good decision.



Mon Feb 13 22:39:35 2006: Subject:   TonyLawrence


*You* have to obey her, but I only have to obey her mother. Therefore you could have followed my lead.. there was no one present to control me, so you could just blame it on me..

------------------------
Comments are closed for this page. You can comment at the "Answer" page.

Take Control of Switching to the Mac

Take Control of VMware Fusion 3

Jump to Comments

Many of the products and books I review are things I purchased for my own use. Some were given to me specifically for the purpose of reviewing them. I resell or can earn commissions from the sale of some of these items. Links within these pages may be affiliate links that pay me for referring you to them. That's mostly insignificant amounts of money; whenever it is not I have made my relationship plain. I also may own stock in companies mentioned here. If you have any question, please do feel free to contact me.

Specific links that take you to pages that allow you to purchase the item I reviewed are very likely to pay me a commission. Many of the books I review were given to me by the publishers specifically for the purpose of writing a review. These gifts and referral fees do not affect my opinions; I often give bad reviews anyway.

We use Google third-party advertising companies to serve ads when you visit our website. These companies may use information (not including your name, address, email address, or telephone number) about your visits to this and other websites in order to provide advertisements about goods and services of interest to you. If you would like more information about this practice and to know your choices about not having this information used by these companies, click here.