Skip to content

Commit acb019d

Browse files
Merge pull request #96 from UCL-ARC/heatherkellyucl-patch-1
Easter closing; Myriad planned outage marked complete
2 parents a66805d + d55aa76 commit acb019d

File tree

1 file changed

+4
-1
lines changed

1 file changed

+4
-1
lines changed

mkdocs-project-dir/docs/Planned_Outages.md

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,8 @@ layout: docs
44

55
# Planned Outages
66

7+
!!! note UCL is closed for Easter from 5pm on Weds 16 to 9am on Weds 23 April
8+
79
The second Tuesday of every month is a maintenance day, when the following clusters should be considered at risk from 8:00AM: Myriad, Kathleen, Young, Michael, Aristotle and the Data Science Platform. We won’t necessarily perform maintenance every month, and notice by email will not always be given about maintenance day work that only puts services at risk.
810

911
Full details of outages are emailed to the cluster-specific user lists.
@@ -16,12 +18,13 @@ After an outage, the first day or two back should be considered 'at risk'; that
1618

1719
Date | Service | Status | Reason
1820
--------------------|---------|--------|--------
19-
7 April 2025 | Myriad | Planned | Outage for switchover to new filesystem. No login access from 9am. Once access is restored later that day or the next day, you will have an empty new home and Scratch with symbolic links to `oldhome` and `oldscratch`. All jobs will be held. You will need to copy your data to the new filesystem or archive it. Once done, you can release the hold on your jobs using `qrls $JOB_ID` or `qrls all`.
21+
2022

2123
## Previous Outages
2224

2325
Date | Service | Status | Reason
2426
--------------------|---------|--------|--------
27+
7 April 2025 | Myriad | Completed | Outage for switchover to new filesystem. No login access from 9am. Once access is restored later that day or the next day, you will have an empty new home and Scratch with symbolic links to `oldhome` and `oldscratch`. All jobs will be held. You will need to copy your data to the new filesystem or archive it. Once done, you can release the hold on your jobs using `qrls $JOB_ID` or `qrls all`. Access restored on 8 April.
2528
18 March 2025 | Myriad, Kathleen | Completed | ACFS will be unavailable and all jobs on Myriad and Kathleen will be drained for 8am. This is to do some work on the ACFS that will allow archiving in future. Had to schedule for 18th not 11th based on vendor availability. Jobs that will not be able to complete before the outage will stay in the queue and will not start until after the outage. The ACFS is at risk all day. We'll let you know if the work is completed earlier.
2629
11 February 2025 | Myriad | Completed | Maintenance day: nodes being drained for a system config change. Also a reboot of ACFS switches will cause the ACFS mount on Myriad to hang for a period during the day.
2730
28 January 2025 | Young, Michael | Completed | /old_lustre will be unmounted. If you still have files in your old home and scratch directories, they will no longer be accessible after this time.

0 commit comments

Comments
 (0)