Merge branch 'master' into report_changes.sh-fix

Author: lvarin

docs/cloud/rahti/tutorials/custom-domain.md

@@ -93,7 +93,7 @@ This is the recommended option to obtain and renew Let's Encrypt certificates. T
 
 1. First, as usual, you need to [install oc](../usage/cli.md#how-to-install-the-oc-tool) and [login into Rahti](../usage/cli.md#how-to-login-with-oc). Then you need to [create a Rahti project](../usage/projects_and_quota.md#creating-a-project). Finally make sure you are in the correct project: `oc project <project_name>`.
 
-1. Double check that the domain name exists. Let's Encrypt needs to verify that you indeed control said domain name, and it does that by issuing a HTTP request to the actual Domain Name and it expects it to respond accordingly. To test it, enter the Domain name in your browser and see that Rahti2 answers accordingly.
+1. Double check that the domain name exists. Let's Encrypt needs to verify that you indeed control said domain name, and it does that by issuing a HTTP request to the actual Domain Name and it expects it to respond accordingly. To test it, enter the Domain name in your browser and see that Rahti answers accordingly.
 
 1. Then you need to create an `Issuer`:
 

docs/computing/systems-puhti.md

@@ -18,7 +18,7 @@ fat tree with a blocking factor of approximately 2:1.
 
 The **Puhti AI** artificial intelligence partition has a total of **80 GPU
 nodes** with a total peak performance of 2,7 petaflops. Each node has
-two latest generation Intel Xeon processors, code name _Cascade Lake_,
+two Intel Xeon processors, code name _Cascade Lake_,
 with 20 cores each running at 2,1 GHz. They also have four Nvidia
 Volta V100 GPUs with 32 GB of memory each. The nodes are equipped with
 384 GB of main memory and 3,6 TB of fast local storage. This partition

docs/computing/usage-policy.md

@@ -53,8 +53,9 @@ suitable disk systems.
 You can use the `csc-workspaces` command to see which cleaning cycle your
 projects are subject to.
 
-A similar procedure will be introduced on Mahti, but it is not yet in place.
-The policy is still that users should keep only actively used data in scratch.
+**Mahti:** A similar procedure will be introduced on Mahti if the disk usage
+grows enough to warrant it. The policy is still that users should keep only
+actively used data in scratch.
 
 ## GPU nodes
 
@@ -106,3 +107,10 @@ If you are still actively using the project you can gain access to the compute
 resources and storage by
 [applying](../accounts/how-to-apply-for-billing-units.md) for more billing
 units.
+
+## Slurm job management by CSC
+
+* CSC will not change job parameters like length or priority. 
+* CSC can terminate jobs if they are misusing resources. E.g., if resources
+  (CPU cores, GPUs, memory) are severely underutilized or IO is overloading
+  the storage system.

docs/data/sensitive-data/sd-access.md

@@ -15,7 +15,7 @@ The steps you need to follow depend on your role in the research team:
 
 ### Project Manager
 
-The project manager is the primary user of a CSC project, responsible for managing the project and project members members. They act as the data controller (or academic organization representative).
+The project manager is the primary user of a CSC project, responsible for managing the project and project members. They act as the data controller (or academic organization representative).
 
 Responsibilities include:
 

docs/support/faq/how-to-resize-in-pouta.md

@@ -6,12 +6,15 @@ It's possible to proceed by taking a snapshot of your instance and then boot the
 You can find more information about snapshot [here](../../cloud/pouta/snapshots.md) and how to proceed.
 
 ### Using the resize functionality 
+
+- [I/O flavors](../../cloud/pouta/vm-flavors-and-billing.md#io-flavors_2) and [GPU flavors](../../cloud/pouta/vm-flavors-and-billing.md#gpu-flavors_2) cannot be resized because data in the root and ephemeral disks are not copied when the resize functionality is used. This means that your machine won't be able to boot and the message "No bootable device" will be displayed in the console. We recommend to use [snapshots](../../cloud/pouta/snapshots.md#launching-a-snapshot-from-the-web-interface).
+
 !!! Warning    
-    [I/O flavors](../../cloud/pouta/vm-flavors-and-billing.md#io-flavors_2) and [GPU flavors](../../cloud/pouta/vm-flavors-and-billing.md#gpu-flavors_2) cannot be resized to a different family flavors. We recommend to use [snapshots](../../cloud/pouta/snapshots.md#launching-an-instance-from-a-volume-snapshot).
+    Keep in mind that when creating a snapshot, only the contents of the root disk will be preserved. If you have data located in the ephemeral disk (usually in the `/mnt` directory), you can back up your data in a [volume](../../cloud/pouta/persistent-volumes.md) or in `Allas` using a tool such as [rclone](../../data/Allas/using_allas/rclone.md) or [s3cmd](../../data/Allas/using_allas/s3_client.md)
 
 !!! Warning  
     It's possible to resize from a `standard` flavor *family* to a `hpc` flavor *family*. Nothing will prevent you to do that but it's **highly not recommended!**  
-    You may lose data during the process and CSC is not responsible. We recommend to only resize to the flavors of the same *family* (except for I/O flavors, see above).
+    You may lose data during the process and CSC is not responsible. We recommend to only resize to the flavors of the same *family*.
 
 In the **Actions** menu of your instance, choose **Resize** to begin the process: