Header Search Widget
IRIS

Instructional & Research Information Systems

Project Storage

Department Project storage

project.eecs short disruption in service at 6:30pm today.

by

project.eecs will have a short disruption of service at 6:30 pm today. This maintenance should last no longer than 15 minutes, and home.eecs should takeover for project during the downtime. Most home directory accesses should recover gracefully from the disruption. Windows mounts may require a re-connection.

Apologies for the short notice. Normally at least 24 hours would be provided before the maintenance window, but this work is required to prevent a more serious problem.
[Read more…] about project.eecs short disruption in service at 6:30pm today.

Performance problems with project storage

by

We have had various reports of slow access times and “out of space” errors when trying to read or write files on project storage. Staff are currently investigating the problem.

If you are experiencing problems, please send the following details to help@eecs:

– name / IP address of client having problems
– description of problem or error message
– type of access -NFS or CIFS?
– name of share you are trying to access

Updates will be posted to this article when we know more.
[Read more…] about Performance problems with project storage

Sutardja Dai Hall power shutdowns

by

Sutardja Dai Hall will have two power shutdowns that impact some storage services. cluster.eecs and inst-fs.eecs will experience a short downtime at the start of each outage, however they will NOT be down for the whole duration of the outage.

The scheduled times for the power outage are as follows:

Thursday July 21st – Power off at 8:00pm and restored at 6am Friday July 22nd

Tuesday July 26th – Power off at 4am and restored at 8am July 26th

Update: the second power shutdown for the 26th was canceled.

Resolved as of 2011-07-26 11:54:00

Storage Migration

by

home.eecs and project.eecs will be down from 8pm to midnight on July
12th, 2011 to be migrated to new equipment. The actual downtime is
expected to be much smaller.

The rest of this notice is primarily targeted at system administrators.
If you choose to read this and have questions about some of the
technical details, please contact your system administrator(s).

1) NFS exports will change. AMD/automaps will be updated accordingly.
Most changes will be straightforward. Some project spaces will be
handled differently, see the details below. Note that this implies
a re-mount.

    Old                        New
home:/eecs/<user>     home:/vol/home/eecs/<user>
project:/proj/<name>  project:/vol/proj/<name>

2) CIFS shares will not change (except for the specially-handled project
spaces). However, SMB2 will be enabled (including durable filehandles).

3) The following project spaces will be migrated differently:

     Old NFS export                 New NFS export
project:/proj/aneureuther       cluster:/vol/cproj/aneureuther
project:/proj/b-cam             cluster:/vol/cproj/b-cam
project:/proj/brass             cluster:/vol/cproj/brass
project:/proj/daw-a             cluster:/vol/cproj/daw-a
project:/proj/demmel-d          cluster:/vol/cproj/demmel-d
project:/proj/graham            cluster:/vol/cproj/graham
project:/proj/hilfinger-a       cluster:/vol/cproj/hilfinger-a
project:/proj/iram              cluster:/vol/cproj/iram
project:/proj/jordan-a          cluster:/vol/cproj/jordan-a
project:/proj/millennium        cluster:/vol/cproj/millennium
project:/proj/neureuther        cluster:/vol/cproj/neureuther
project:/proj/neureuther-flcc   cluster:/vol/cproj/neureuther-flcc
project:/proj/neureuther-intel  cluster:/vol/cproj/neureuther-intel
project:/proj/now               cluster:/vol/cproj/now
project:/proj/parlab            cluster:/vol/cproj/parlab
project:/proj/radlab            cluster:/vol/cproj/radlab
project:/proj/tinyos            cluster:/vol/cproj/tinyos
project:/proj/titanium          cluster:/vol/cproj/titanium
project:/proj/vision            cluster:/vol/cproj/vision
project:/proj/vision-b          cluster:/vol/cproj/vision-b
project:/proj/vision-c          cluster:/vol/cproj/vision-c
project:/proj/vision-e          cluster:/vol/cproj/vision-e
project:/proj/yelick            cluster:/vol/cproj/yelick
project:/proj/yelick-old        cluster:/vol/cproj/yelick-old

If the project space in the above list had CIFS share(s) associated with
it, the said share(s) will retain their name, but will now live on
\\\\\\CLUSTER\<share_name>.

If you have questions or concerns about the timing or technical details,
then please email me at aculich@eecs.berkeley.edu or drop by 319 Soda;
as many of you know, Fred Archibald has retired and I have taken over
his role managing IDSG, so whether or not you have storage questions,
feel free to drop by my office to say hello and introduce yourself.
[Read more…] about Storage Migration

Unplanned SAN outage

by

An unplanned outage of our Storage Area Network (SAN) was quickly addressed by staff. The outage affected home.eecs and project.eecs storage, as well as the storage used by imap.eecs.

Staff responded to automated alerts, and were able to bring the SAN back online. The precise cause of the outage currently remains unknown.

Resolved as of 2011-05-09 21:05:00