Skip to content

Commit b496691

Browse files
committed
hpc-cache: add nas prereqs & note about update
1 parent 9dd9d44 commit b496691

File tree

2 files changed

+69
-11
lines changed

2 files changed

+69
-11
lines changed

articles/hpc-cache/hpc-cache-manage.md

Lines changed: 24 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -4,27 +4,44 @@ description: How to manage and update Azure HPC Cache using the Azure portal
44
author: ekpgh
55
ms.service: hpc-cache
66
ms.topic: conceptual
7-
ms.date: 1/08/2020
7+
ms.date: 1/29/2020
88
ms.author: rohogue
99
---
1010

1111
# Manage your cache from the Azure portal
1212

13-
The cache overview page in the Azure portal shows project details, cache status, and basic statistics for your cache. It also has controls to delete the cache, flush data to long-term storage, or update software.
13+
The cache overview page in the Azure portal shows project details, cache status, and basic statistics for your cache. It also has controls to stop or start the cache, delete the cache, flush data to long-term storage, and update software.
1414

1515
To open the overview page, select your cache resource in the Azure portal. For example, load the **All resources** page and click the cache name.
1616

1717
![screenshot of an Azure HPC Cache instance's Overview page](media/hpc-cache-overview.png)
1818

1919
The buttons at the top of the page can help you manage the cache:
2020

21+
* **Start** and [**Stop**](#stop-the-cache) - Suspends cache operation
2122
* [**Flush**](#flush-cached-data) - Writes changed data to storage targets
2223
* [**Upgrade**](#upgrade-cache-software) - Updates the cache software
2324
* **Refresh** - Reloads the overview page
2425
* [**Delete**](#delete-the-cache) - Permanently destroys the cache
2526

2627
Read more about these options below.
2728

29+
## Stop the cache
30+
31+
You can stop the cache to reduce costs during an inactive period. You are not charged for uptime while the cache is stopped, but you are charged for the cache's allocated disk storage. (See the [pricing](https://aka.ms/hpc-cache-pricing) page for details.)
32+
33+
A stopped cache does not respond to client requests. You should unmount clients before stopping the cache.
34+
35+
The **Stop** button suspends an active cache. The **Stop** button is available when a cache's status is **Healthy** or **Degraded**.
36+
37+
![screenshot of the top buttons with Stop highlighted and a pop-up message describing the stop action and asking 'do you want to continue?' with Yes (default) and No buttons](media/stop-cache.png)
38+
39+
After you click Yes to confirm stopping the cache, the cache automatically flushes its contents to the storage targets. This process might take some time, but it ensures data consistency. Finally, the cache status changes to **Stopped**.
40+
41+
To reactivate a stopped cache, click the **Start** button. No confirmation is needed.
42+
43+
![screenshot of the top buttons with Start highlighted](media/start-cache.png)
44+
2845
## Flush cached data
2946

3047
The **Flush** button on the overview page tells the cache to immediately write all changed data that is stored in the cache to the back-end storage targets. The cache routinely saves data to the storage targets, so it's not necessary to do this manually unless you want to make sure the back-end storage system is up to date. For example, you might use **Flush** before taking a storage snapshot or checking the data set size.
@@ -52,6 +69,8 @@ The software update can take several hours. Caches configured with higher throug
5269

5370
When a software upgrade is available, you will have a week or so to apply it manually. The end date is listed in the upgrade message. If you don't upgrade during that time, Azure automatically applies the update to your cache. The timing of the automatic upgrade is not configurable. If you are concerned about the cache performance impact, you should upgrade the software yourself before the time period expires.
5471

72+
If your cache is stopped when the end date passes, the cache will automatically upgrade software the next time it is started. (The update might not start immediately, but it will start in the first hour.)
73+
5574
Click the **Upgrade** button to begin the software update. The cache status changes to **Upgrading** until the operation completes.
5675

5776
## Delete the cache
@@ -63,13 +82,14 @@ The back-end storage volumes used as storage targets are unaffected when you del
6382
> [!NOTE]
6483
> Azure HPC Cache does not automatically write changed data from the cache to the back-end storage systems before deleting the cache.
6584
>
66-
> To make sure that all data in the cache has been written to long-term storage, follow this procedure:
85+
> To make sure that all data in the cache has been written to long-term storage, [stop the cache](#stop-the-cache) before you delete it. Make sure that it shows the status **Stopped** before clicking the delete button.
86+
<!--... written to long-term storage, follow this procedure:
6787
>
6888
> 1. [Remove](hpc-cache-edit-storage.md#remove-a-storage-target) each storage target from the Azure HPC Cache by using the delete button on the Storage targets page. The system automatically writes any changed data from the cache to the back-end storage system before removing the target.
6989
> 1. Wait for the storage target to be completely removed. The process can take an hour or longer if there is a lot of data to write from the cache. When it is done, a portal notification says that the delete operation was successful, and the storage target disappears from the list.
7090
> 1. After all affected storage targets have been deleted, it is safe to delete the cache.
7191
>
72-
> Alternatively, you can use the [flush](#flush-cached-data) option to save cached data, but there is a small risk of losing work if a client writes a change to the cache after the flush completes but before the cache instance is destroyed.
92+
> Alternatively, you can use the [flush](#flush-cached-data) option to save cached data, but there is a small risk of losing work if a client writes a change to the cache after the flush completes but before the cache instance is destroyed.-->
7393

7494
## Cache metrics and monitoring
7595

articles/hpc-cache/hpc-cache-prereqs.md

Lines changed: 45 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -4,7 +4,7 @@ description: Prerequisites for using Azure HPC Cache
44
author: ekpgh
55
ms.service: hpc-cache
66
ms.topic: conceptual
7-
ms.date: 10/30/2019
7+
ms.date: 02/12/2020
88
ms.author: rohogue
99
---
1010

@@ -65,12 +65,6 @@ The cache supports Azure Blob containers or NFS hardware storage exports. Add st
6565

6666
Each storage type has specific prerequisites.
6767

68-
### NFS storage requirements
69-
70-
If using on-premises hardware storage, the cache needs to have high-bandwidth network access to the datacenter from its subnet. [ExpressRoute](https://docs.microsoft.com/azure/expressroute/) or similar access is recommended.
71-
72-
NFS back-end storage must be a compatible hardware/software platform. Contact the Azure HPC Cache team for details.
73-
7468
### Blob storage requirements
7569

7670
If you want to use Azure Blob storage with your cache, you need a compatible storage account and either an empty Blob container or a container that is populated with Azure HPC Cache formatted data as described in [Move data to Azure Blob storage](hpc-cache-ingest.md).
@@ -89,6 +83,50 @@ It's a good practice to use a storage account in the same location as your cache
8983

9084
You also must give the cache application access to your Azure storage account as mentioned in [Permissions](#permissions), above. Follow the procedure in [Add storage targets](hpc-cache-add-storage.md#add-the-access-control-roles-to-your-account) to give the cache the required access roles. If you are not the storage account owner, have the owner do this step.
9185

86+
### NFS storage requirements
87+
88+
If using an NFS storage system (for example, an on-premises hardware NAS system), make sure it meets these requirements. You might need to work with the network administrators for your storage system or data center to verify these settings.
89+
90+
> [!NOTE]
91+
> Storage target creation will fail if the cache has insufficient access to the NFS storage system.
92+
93+
* Network connectivity: The Azure HPC Cache needs high-bandwidth network access between the cache subnet and the NFS system's data center. [ExpressRoute](https://docs.microsoft.com/azure/expressroute/) or similar access is recommended.
94+
95+
* Port access: The cache needs access to specific TCP/UDP ports on your storage system. Different types of storage have different port requirements.
96+
97+
To check your storage system's settings, follow this procedure.
98+
99+
* Issue an `rpcinfo` command to your storage system to check the needed ports. The command below lists the ports and formats the relevant results in a table. (Use your system's IP address in place of the *<storage_IP>* term.)
100+
101+
```bash
102+
rpcinfo -p <storage_IP> |egrep "100000\s+4\s+tcp|100005\s+3\s+tcp|100003\s+3\s+tcp|100024\s+1\s+tcp|100021\s+4\s+tcp"| awk '{print $4 "/" $3 " " $5}'|column -t
103+
```
104+
105+
* Ensure that these ports on your storage system allow inbound and outbound traffic:
106+
107+
| Protocol | Port | Service |
108+
|----------|-------|----------|
109+
| TCP/UDP | 111 | rcpbind |
110+
| TCP/UDP | 2049 | NFS |
111+
| TCP/UDP | 4045 | nlockmgr |
112+
| TCP/UDP | 4046 | mountd |
113+
| TCP/UDP | 4047 | status |
114+
115+
* Check firewall settings to be sure that they allow traffic on all of these required ports. Be sure to check firewalls used in Azure as well as on-premises firewalls in your data center.
116+
117+
* Directory access: Enable the `showmount` command on the storage system. Azure HPC Cache uses this command to check that your storage target configuration points to a valid export, and also to make sure that multiple mounts don't access the same subdirectories (which risks file collisions).
118+
119+
> [!NOTE]
120+
> If your NFS storage system uses NetApp's ONTAP 9.2 operating system, **do not enable `showmount`**. [Contact Microsoft Service and Support](hpc-cache-support-ticket.md) for help.
121+
122+
* Root access: The cache connects to the back-end system as root (with the user ID 0). Check these settings on your storage system:
123+
124+
* Enable `no_root_squash`. This option ensures that the remote root user can access files owned by root.
125+
126+
* Check export policies. For example, make sure that root access is not restricted to internal IP addresses.
127+
128+
* NFS back-end storage must be a compatible hardware/software platform. Contact the Azure HPC Cache team for details.
129+
92130
## Next steps
93131

94132
* [Create an Azure HPC Cache instance](hpc-cache-create.md) from the Azure portal

0 commit comments

Comments
 (0)