What is going on with UPCP??????

[mr.wonderful]

Everyone of our boxes is having issues updating via /scripts/upcp. WGET failes to complete, have to keep hitting cntrl-c to continue and the updates gets stuck. We get booted out of SSH2. Its not just one box but every single one of them. Whats going on with the updates?

In the contol panel for user sites we are seeing this.

There was an error setting uids. Please have the admin check permissions on: /usr/local/cpanel/bin/mysqlwrap CP-Wrap Critcal Error (signal 13 from wrapped program)! This may indicate a corrupt admin binary in /usr/local/cpanel/bin/

 

[mr.wonderful]

Originally posted by thaphantom
working fine from here, check your network

Network? You got to be kidding, Nothing to do with network.

 

[PWSowner]

I just updated an hour ago with no problem.

 

[65535]

Check that your DNS server in /etc/resolv.conf is responding

Rus

 

[icanectc]

try disabling firewall we had this problem when running APF. If one is installed.

 

[kris1351]

Had a server doing this last week. Do a ps -ef | grep upcp and ps -ef | grep up2date. We had a upcp get hung and it was because there was already an up2date was running and then both were stuck.

 

[Website Rob]

Can we get some stats on what people are upgrading.

WHM 9.3.0 cPanel 9.3.0-E65
RedHat 9 - WHM X v3.1.0

upgrading to 9.3.0 (build 105) and got hung up here:

cPanel Layer 2 Update Commencing
==> Installing.......
==> Running post install scripts....sync
syncing cpanel binaries
syncing perl libraries
syncing web host manager
syncing source files


Top does show: /usr/bin/perl /scripts/cpanelsync httpupdate
as running so it could just be taking a long. Not sure if because of upgrading from 65 to 105 requiring a lot of changes or not. Did do a ps -ef | grep upcp and ps -ef | grep up2date and no problem there.


Quick update, while the update had sat at the above mentioned part of the update for 10 minutes before making this post, during the making of this post the upgrade had progressed another 4 lines further:

syncing 3rdparty programs
syncing cgi-sys
syncing java-sys
syncing web based 3rdparty programs

Still sitting there now but Cpanel index page is showing:

MySQL Databases - There was an error setting uids. Please have the admin check permissions on: /usr/local/cpanel/bin/mysqlwrap / 25

and I'm hoping the update is just taking a long time and once done, will self-correct the mySQL errors.

 

[Website Rob]

'nother quick update.

Seems it never went past where it was stuck as upcp and httpupdate went to Sleep. After killing the upcp process (use ps aux|grep upcp) to find it, and running the update again, all is well once more -- and now upgraded to E-105.

oops - forgot to mention you have to kill httpupdate as well. And cpsrvd failed on me but no big deal to restart it.

 

[mr.wonderful]

I dont have any of the above processes running and i can assure you the next time i run upcp it will hang. Whats up with this? It needs some serious fixing.

 

[tAzMaNiAc]

Originally posted by mr.wonderful
I dont have any of the above processes running and i can assure you the next time i run upcp it will hang. Whats up with this? It needs some serious fixing.

Did you try ...... tracerouting to layer1.cpanel.net ? If all your servers are on the same network (bad idea IMHO) and all of them have the same problem, it could be a routing/network issue.

So, that is why I am asking if you did:

traceroute layer1.cpanel.net

yet?

Brenden

 

[mr.wonderful]

Originally posted by tAzMaNiAc
Did you try ...... tracerouting to layer1.cpanel.net ? If all your servers are on the same network (bad idea IMHO) and all of them have the same problem, it could be a routing/network issue.

So, that is why I am asking if you did:

traceroute layer1.cpanel.net

yet?

Brenden

The process its getting stuck on is

/usr/bin/perl /scripts/cpanelsync httpupdate.cpanel.net /cpan

Then i have to log into another session kill it off then it tells me 100% done and continues the update. All my servers are on different networks so the traceroute seemed fine.

 

[PWSowner]

I just upgraded again from 9.3 E64 to E116. I got a longer than normal delay at syncing source files, but then it completed just fine.

 

[nettigritty]

getting a hung upcp and easyapache !

 

[magin]

same here

tried to rebuild rpm db by shell and got stuck, the same with upcp

 

[nettigritty]

an RPM rebuild did it for me - working fine now

 

[magin]

i red i cpanel faq that this can be caused cause tmp is running suexec,

need somebody tell me what are the defaults permissions for /tmp directory 777?

thanks

 

[nettigritty]

yes 777