HPE Primera

HPE Discover Virtual – Storage News

IT Conference season is almost upon us but this time with a difference in that they have gone online. HPE Discover Virtual is about to kick off in a couple off weeks and an advantage over the physical show is that its free. HPE have shared their storage news before the show, the news focuses on the SAN systems Nimble and Primera.

Primera

Peer Persistence and Replication

Peer Persistence was always a popular feature with 3PAR and one of most read posts on my blog. With good reason, it gave you a Metro cluster for increased availability and flexibility across data centres. HPE are now announcing Peer Persistence is coming to Primera. I am waiting to see a demo but HPE advise it is is easy to setup and will be included free with Primera. Replication capabilities are also enhanced with near-instant asynchronous replication over extended distances, this allows an RPO down to one minute.

Hardware

In terms of hardware Primera is now capable of supporting all NVMe drives.

AI

To ensure optimum performance Primera uses machine learning gathered both locally and from the entire install base to make resource allocation decisions .

Nimble

Nimble has been able to look inside your VMware environment for some time and make performance recommendations with it’s cross-stack analytics. That feature is now also available to those using Hyper-V. You can take a look at how this looked for VMware uses to get a feel for how this will function.

Nimble data management features also get an upgrade with the addition of replication to a 3rd site. The third site could for example be another Nimble array in a DR site or potentially a cloud location by utilising HPE Cloud Volumes. We previously covered Cloud Volumes in detail if you want a refresher.

Nimble also gets a hardware boost by adding support for storage class memory. Storage class memory sits in between SSD and memory offering a blend of enhanced performance at a median cost point. This SCM can be used as it is in the Nimble systems to expand the capacity of cache, benefiting the performance by getting more cache hits and reducing the load on SSD’s

I will hopefully be able to bring you more details after the HPE Virtual event.

Further Info

Video summarising the announcements

Podcast of the announcements

Blocks and Files blog post

June 11, 2020June 11, 2020

Richard Arnold1 Comment

HPE 3PAR & Primera SSMC Topology Insights

Hello, starting with SSMC 3.6, HPE introduced a new feature called topology Insights.

Let me introduce what topology Insights is and how to enable it.

This guest post is brought to you by Armin Kerl, if you fancy trying you hand at blogging check out our guest posting opportunities.

What is SSMC Topology Insights?

Topology Insights allows a consolidated view of both storage and the VMware layer. To see it within SSMC go to: VMWARE > Virtual Machines

Change to: Topology view from the drop down in the right hand window

You will see all VM’s and the relationship between:

VM <> VMDK <> Datastore <> ESXi Server <> Virtual Volume <> Storage

Also, any performance impacts are displayed on top with a selectable time line, red shows a problem.

How to get Topology Insights?

You need at least:

SSMC 3.6
3PAR with OS 3.3.x
Service Processor 5.x
VMware vCenter (no Hyper-V currently)

In the SSMC go to Settings

In Section “Application” enable “Topology”

and VMware > Virtual machines

Now go to “Storage Systems” > “Systems” > “Actions”

Enter the “Set SP credentials” here.

If it is successful, check under “Systems” > “Settings” if the vCenters entry is here.

If not, go to edit and enter here.

If you don’t see “Set SP credentials”?

Like this:

Or get this Message?

I have had several customer sites, where the entry did not appear.

Most of the time these are storage systems and SP, which have been upgraded from 3.2.x to 3.3.x.

The following is for information only, it is recommended to contact support before proceeding.

It is a little bit Tricky, but here is a way to fix it:

Enter: cli% getsys -showsysobjs

Look for {ServiceProcessorCookie {}}

Wrong: … {SessionTimeout 10800} {ServiceProcessorCookie {}} …

Right: … {SessionTimeout 10800} {ServiceProcessorCookie SPxxxyyyyzzzz:IP.of.the.SP} …

If the SP Cookie is empty “{}”, we need to fix it.

Here we see the SPIP = IP address and SPID = SP ID

Now set the missing Values:

cli% setsys ServiceProcessorCookie SPID:SPIP

Example: setsys ServiceProcessorCookie SPCZ123456789:192.168.10.23

Check again: Enter: cli% getsys -showsysobjs

Now OK?

Go back to the SSMC and the “Set SP credential” should appear.
This fix worked for me every Time.

Thank you, folks for reading.

Armin

December 10, 2019December 10, 2019

Richard ArnoldLeave a comment

HPE Discover – Multiple new storage systems unveiled

Today at HPE Discover a large number of storage announcements where made, including some new product lines.

This is a brand new storage system from HPE aimed at the high end. I asked the product manager about the heritage of the system, if it was 3PAR or Nimble based and they advised that it was a new from the ground up system combining the best of both. Primera starts in a two node configuration available as all flash and hybrid, it can scale to 4 nodes capable of delivering 1.5 million IOPs. The influence from the 3PAR and Nimble family is obvious in the feature set and key capabilities being marketed.

Like 3PAR today the Primera system is based on an active active design with all data striped across all components in order to maximise performance and resiliency.

InfoSight which was originally developed by Nimble is used extensively in conjunction with Primera. InfoSight is described by HPE as providing global intelligence “Powered with the most advanced AI for infrastructure”. This is essentially describing that InfoSight will be gathering data from the entire install base and using machine learning to identify any potential issues or improvements and inform users. HPE estimate that 90% of issues are non-storage related, InfoSight is able to discover potential issues beyond the storage layer to enable a complete wrap around to infrastructure services.

Primera is powered by a modular OS which means that software components of the system are decoup. Today planning an upgrade to a storage system is a major undertaking with an all or nothing decision of either do or don’t upgrade. The new modular design has decoupled data services allowing individual services to be updated. This new design should allow safer updates with fewer reboots and allow a more agile route to updates and deployment which should in turn enhance innovation and willingness to apply updates.

Possibly the most surprising aspect of the new system is that HPE are guaranteeing 100% availability. This guarantee is available on all Primera systems without qualification, it seems this is enabled not only by HPE’s confidence in the new system but again because of the Intelligence that InfoSight will give them. This will ensure that customers systems are configured are correctly , any issues are remediated before they occur and if appropriate fixes rolled out to other customers. Primera has been developed to be simple even though it is an advanced high end system with time to deployment estimated at 20 minutes.

SimpliVity Updates

The SimpliVity hyper-converged platform received some new models and enhancements also

InfoSight – Now works with SimpliVity. Similar to other implementations this make proactive health recommendations and other insights beyond the storage layer for example into noisy VMs
Archive node – This is a new 2U hardware offering that contains a mixture of HD and SSD disks. The node is intended as the name suggests for longer term storage
SimpliVity 325 – This is a new model in the range that takes on a 1U form factor. This dense format is intended for environments where space is at a premium

Nimble Storage dHCI

This is again a brand new system from HPE, it is a Hyper-converged product that enables you to scale storage and compute separately. The system is designed for simplicity in 5 clicks the entire stack is setup and ready for vCenter.

Simplicity of HCI flexibility of converged

I think this latest batch of updates show a significant engineering effort and understanding towards the requirements of todays datacentre customers who want flexibility, simplicity and performance.

Stay tuned for more information on the new announcements. Any questions let me know and I will try to find out the answer.

Tag: HPE Primera