Header Search Widget

IRIS

Instructional & Research Information Systems

LDAP & Jabber outage this morning, May 15, 2013

by

The department’s primary LDAP server ldap.eecs stopped responding early this morning around 12:30am. This affected a few other services which rely on ldap.eecs; we had reports of problems accessing Repo, and jabber.eecs stopped accepting connections at some point.

ldap.eecs was restarted at 7:00am and this would have restored most affected services. However, the jabber server seemed to be hung and had to be restarted; this was not noticed until around 10:30am.

Our ldap servers are aging and there is a current project underway to upgrade them to more modern hardware and software. An announcement about the scope of this project will be made in the coming weeks.

We are also planning to implement better monitoring of our jabber server so that future outages will be detected much more quickly.

Resolved as of 2013-05-15 07:00:00

May 2013 Server Patching

by

Microsoft will release their patches for the month on Tuesday, May 14th, 2013.

PRINT and WINTERM will automatically patch and reboot Wednesday, May 15th, 2013, at 3:00 am.

WINTERM will be unavailable till 10:00am May 15th, 2013.

Resolved as of 2013-05-17 11:43:00

May 3, 2013 directory server malfunctioning

by

LDAP directory service from ldap.CS.Berkeley.EDU was malfunctioning from about Friday May 3, 2013 01:56 to about 06:57.

It appears authenticated outgoing smtp service on gateway.EECS.Berkeley.EDU also malfunctioned during that time frame when gateway.EECS.Berkeley.EDU failed to use ldap.EECS.Berkeley.EDU for LDAP directory service instead of ldap.CS.Berkeley.EDU

Resolved as of 2013-05-03 06:57:00

Mailing list mail delayed over the weekend

by

The department’s mailing list server ran out of disk space late Friday, April 26, around 4:30pm. A single very large message to one list caused what should have been a temporary issue. I removed some unnecessary files on Friday and thought that everything was working normally, but this morning noticed messages were not being distributed.

The server was restarted and has since caught up with all outstanding messages. Because of this you may notice messages sent over the last 3 days just showing up in your mailbox today, but no mail has been lost.

An investigation is underway as to why our old monitoring software did not alert us to this problem sooner. A new monitoring solution is being deployed in IRIS and this server will be moved to this new monitoring solution as soon as possible.
[Read more…] about Mailing list mail delayed over the weekend

Partial loss Wireless Network on 4th floor of Soda hall

by

Part of wireless network on the 4th floor of Soda hall was down from 10:42am to 10:53am because that UPS failed during the power UPS testing from the electrician. Those APs are back on around 10:53am. Sorry for the inconvenience that this may cause. If you still experience any wireless issue in that area, please report to help@eecs

Resolved as of 2013-04-26 10:53:00

Retiring web redirects on www.eecs.berkeley.edu

by

IRIS offers a redirect service for our account holders to forward their http://www.eecs.berkeley.edu/~username (and www.cs.berkeley.edu) web page to another location. This is typically used when an account is closed.

We promise this forwarding for one year after an account is closed, but have never done a bulk cleanup, and the list has grown quite large. As a first step to cleaning these up, we are now checking the destination URLs to make sure they are still valid. On 9am on April 30, 2013, the attached list of URLs will be removed from the server configuration.

If you have an update for any of these redirects, please contact the IRIS helpdesk.

[Note: In the attached list, the URL we are redirecting is on the left side, the destination is on the right, and the HTTP response code is listed, or -1 if the remote host cannot be found or does not answer.]

Attachment: bad-URLs.txt
[Read more…] about Retiring web redirects on www.eecs.berkeley.edu