freeBSD & Current

[ckh]

Upgraded to the R118 recently and started running into problems.

1. update ports option is turned off, but, it insists on updating it anyway. Really slows down the server and install process as it seems to delete the index-5 file and then tries to rebuild it. If I do a fetchindex during the install it works better

2. Lost my 'Show Current CPU Usage'. It would display the header, but just not show any of the current running processes, similar to the linux problem. Tried doing a upcp --force and that didn't help. Put in a ticket and it went to level II, but, they said:

When I strace it I see:

First a bunch of:
[pid 74887] stat(\"/etc/nsswitch.conf\", 0xbfbff750) = -1 ENOENT (No such file or directory)
[pid 74887] stat(\"/etc/nsswitch.conf\", 0xbfbff750) = -1 ENOENT (No such file or directory)
[pid 74887] stat(\"/etc/nsswitch.conf\", 0xbfbff750) = -1 ENOENT (No such file or directory)
[pid 74887] stat(\"/etc/nsswitch.conf\", 0xbfbff750) = -1 ENOENT (No such file or directory)
[pid 74887] stat(\"/etc/nsswitch.conf\", 0xbfbff750) = -1 ENOENT (No such file or directory)

then it opens etc/spwd.db and seeks in it for a while and then opens it and seeks ... over and over and over...

So I attempted a upcp and it jusy hangs trygin to do the ports... (I had to control C each one so it would continue)

Its running very very slow and hasn't conpleted yet and its been hours. (I finally ended it by control-C-ing my way out.

I imagine that WHM->System Health->Show Current CPU Usage is timing out waiting for top to finish.

Likely this is due to a firewall issue or whatever is causing upc to run so slow.

I imagine it may have something to do with whatever is causing /etc/nsswitch.conf and /etc/spwd.db to loop when I strace
WHM->System Health->Show Current CPU Usage

It was running slowly as the index-5 file got deleted again and it was rebuilding it. I did the --force thing and it didn't change anything.

So, today I went to S14. Started updating the ports again and deleted the index-5 file and brought the server to a crawl. I did a fetchindex during the installation and that helped.

After the install, my "Show Current CPU Usage" returned to normal and everything else seems to work.

Anyone else with freeBSD experience this? I saw the thread where a lot of linux boxes lost their show cpu usage, but didn't see any BSD boxes.

[guinea812]

Yeah, both the FreeBSD machines I have exhibit all the items that you mentioned. I hope they fix thoes problems soon.

[ckh]

You may want to open a support ticket. They couldn't fix it and said it was a problem with my server. When I told them I went to stable and it fixed everything, they said it must be an isolated incident as they didn't have any problems on their servers....

Chris

[ckh]

What version of freeBSD are you running? I'm running 5.1 and support was asking me if I knew what version you were running.

My ticket number is 80321.

Chris

[guinea812]

Hmmmm, I should open a ticket then. I am running one on FreeBSD 4.8 and one on FreeBSD 4.9.

[ckh]

When it happened, did were you upgrading to release? Did you go back to stable and if so did it resolve it?

I got a support ticket in and they are asking these questions. Did you put in a ticket also??

Thanks
Chris

[eos1]

2. Lost my 'Show Current CPU Usage'. It would display the header, but just not show any of the current running processes

Same Here on FreeBSD 4.9

[eos1]

2. Lost my 'Show Current CPU Usage'. It would display the header, but just not show any of the current running processes

Just upcp one of FreeBSD 5.2.1 to:
WHM 9.9.8 cPanel 9.9.8-E128
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0
and Lost 'Show Current CPU Usage'.

upcp to:
WHM 9.9.8 cPanel 9.9.8-S14
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0

I got back 'Show Current CPU Usage'

[eos1]

Following is NOT ready...

FreeBSD

RELEASE
9.9.8-RELEASE_118
(Fri Nov 12 11:15:08 2004)

CURRENT
9.9.8-CURRENT_117
(Fri Nov 12 09:38:18 2004)

EDGE
9.9.8-EDGE_128
(Sat Nov 20 06:51:18 2004)

Please fix this, Nick...

[ckh]

I take this to mean that release, current, and edge all have this problem and that going back to stable fixed it??

Thanks
Chris

[eos1]

I take this to mean that release, current, and edge all have this problem and that going back to stable fixed it??

Thanks
Chris

You have problem with 9.9.8-RELEASE_118
I had problems with 9.9.8-CURRENT_117(FreeBSD 4.9) and 9.9.8-EDGE_128(FreeBSD 5.2.1)

= not ready.

I didn't put back 9.9.8-CURRENT_117(FreeBSD 4.9) to 9.9.8-S14, but I'm sure upcp to 9.9.8-S14 will solve this too like my upcp 9.9.8-EDGE_128 to 9.9.8-S14.

[eos1]

Just upcp one of FreeBSD 5.2.1 to:
WHM 9.9.8 cPanel 9.9.8-E128
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0
and Lost 'Show Current CPU Usage'.

upcp to:
WHM 9.9.8 cPanel 9.9.8-S14
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0

I got back 'Show Current CPU Usage'

again, upcp to:
WHM 9.9.8 cPanel 9.9.8-E128
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0
and Lost 'Show Current CPU Usage' again.

/scripts/upcp --force to:
WHM 9.9.8 cPanel 9.9.8-E128
FreeBSD 5.2.1-RELEASE i386 - WHM X v3.1.0
Never get back 'Show Current CPU Usage'

I think you need to wait for next builds.

[ckh]

http://bugzilla.cpanel.net/show_bug.cgi?id=1664

Please add your comments so we can get rid of this bug!!

Chris

[eos1]

Fixed in 9.9.8-E155.
Confirmed!

Thanks, Nick!!!