The 'I work in I.T' thread

[Donny_Dog]

Never known anything like it with Broadcom/VMware. It's a proper s**t show.

 

[.Joe]

Never known anything like it with Broadcom/VMware. It's a proper s**t show.

Messagelabs/Broadcom experience gave me a short warning but still managed to make a mess just as finally got stable infra

 

[rctempire]

Never known anything like it with Broadcom/VMware. It's a proper s**t show.

I dont see the logic at all, carve out the whole profitable business, cut everyone out and go direct. Makes sense, but its an awful mess.

 

[Advikaz]

Mimecast.






Just Mimecast. It's like they don't want to sell any solutions.

 

[ChrisR]

Mimecast.






Just Mimecast. It's like they don't want to sell any solutions.

I know the new CEO there, I'll let him know 😂

 

[Advikaz]

I know the new CEO there, I'll let him know 😂

Honestly a nightmare to deal with.

The product itself is pretty good but their whole model of customer service, support and the sales process is horrendously slow and painful.

 

[boultonn]

Mimecast.






Just Mimecast. It's like they don't want to sell any solutions.

Consulting for a client who use the mimecast agent and it must have tripled how long it took to get the laptop built, even a service desk bod who deploys these every day having to go through a whole playbook just to get it deployed and working.
(Don't get me started on the targeted threat protection).

 

[rctempire]

Mimecast.






Just Mimecast. It's like they don't want to sell any solutions.

Funny i had to deal with them last week for this exchange stuff. Imagine not allowing users to add office365. They have to do it manually on the background and then it shows as a tick box.
Weird.

 

[Advikaz]

Yeah agreed. It's frustrating as said as the solution isn't bad generally.

But dealing with them is the most frustrating supplier I think i've had to deal with on a regular basis. Even pricing it out takes weeks as you have to get an individual f**king quote from them every time that's client specific (even though internally they run on <49 users <100 users models etc.

Their pricing model makes me want to headbutt a wall.

Literally no need for it to be this awkward.

FML!

 

[rctempire]

Today i ran into GPO madness.
This business has 500+ GPOs.
Its brilliant trying to find the one interrupting what is going on. Even RSOP and GPO Result is saying "Good luck."

 

[Krarl]

 

[andybond]

Today i ran into GPO madness.
This business has 500+ GPOs.
Its brilliant trying to find the one interrupting what is going on. Even RSOP and GPO Result is saying "Good luck."

New OU.

Block inhertiance
Add policies in until you find the policy that over wrote the previous setting

Suggest post working out that a monolithic policy is created for items that stay static and a fast change policy is deployed for things that change , such as allowed lists for example.

I’d also suggest running GPResult and login to a file to see what is being processed.

 

[Gus]

I thought I'd throw this out there.

Anyone living near Harlow interested in a data centre engineer role for a finance company? Give me a shout on pm.

 

[rctempire]

New OU.

Block inhertiance
Add policies in until you find the policy that over wrote the previous setting

Suggest post working out that a monolithic policy is created for items that stay static and a fast change policy is deployed for things that change , such as allowed lists for example.

I’d also suggest running GPResult and login to a file to see what is being processed.

Plan is to scrap 499 of them with a move to windows 11.
So its fresh and easily rationalised.


Either way, my Citrix implementation is going ahead Monday - Friday rollout of 1000+ users.
Hurrah. Azure madness ensues.

 

[andybond]

Plan is to scrap 499 of them with a move to windows 11.
So its fresh and easily rationalised.


Either way, my Citrix implementation is going ahead Monday - Friday rollout of 1000+ users.
Hurrah. Azure madness ensues.

That’s what I was suggesting.

I have done extensive research and performance monitoring on monolithic vs dynamic GPO.

GO ME! 👀

 

[charltjr]

I'm currently redeploying a customer's production cloud environment for an upgrade.

a) I left this stuff behind years ago for a reason, it's boring yet stressful
b) It's PAINFUL. Ancient product stack (OpsWorks ) and manual actions all over the place

I'm therefore posting on CS while waiting for CodePipeline to finish running it's s**t. Really looking forward to the hour or so of waiting for database update scripts to finish. Oh what fun.

Roll on the new product stack with actual automation.

 

[mace¬]

Thought we may have had some sort of DDOS today so had my first look at what the SOC do.

So disappointed they never brought up a world map with loads of red pings like some nuclear war thing to see where the traffic was coming from. Properly let down.

 

[SharkyUK]

N00bs.

I'm rewriting the deployment, tracking and guidance software for some submarine-carried missile or other. The previous guy just got sacked.

 

[mace¬]

If I send you a list can you hardcode house numbers and postcodes?

 

[.Joe]

Thought we may have had some sort of DDOS today so had my first look at what the SOC do.

So disappointed they never brought up a world map with loads of red pings like some nuclear war thing to see where the traffic was coming from. Properly let down.

https://threatmap.checkpoint.com/

We have this up on a wall board for when the odd customer visits to make it look like we actually do work

 

[Advikaz]

If I send you a list can you hardcode house numbers and postcodes?

Fingers crossed for Bradford 😍

 

[SharkyUK]

Fingers crossed for Bradford 😍

GIGO

 

[Krarl]

Fingers crossed for Bradford 😍

Oi

 

[Donny_Dog]

Went to a Microsoft consultation in the village next to me today (Eggborough -north Yorks). They're putting a data centre on the old power station site.
Be done for 2027 apparently.
Dedicated to Azure & AI.

Asked if it was going to be 'azure north' given south and west (London and Cardiff) already in place - but they just said, it hasn't been confirmed.

I'll be fancying a job there.
"Canteen staff wanted" sweet

 

[rctempire]

N00bs.

I'm rewriting the deployment, tracking and guidance software for some submarine-carried missile or other. The previous guy just got sacked.

I hope its related to Trident. They have had a bad week......

 

[Gray]

Recently configured and deployed my first VMware vSAN cluster... VMotion'd all the production VM's over from the legacy hosts last night - No complaints, nor did anybody on-site notice

Got the process nailed down, encountered a weird issue with multiple datastores being created - Logged a ticket with VMware, two days later they just said "delete them" ha.

 

[andybond]

Recently configured and deployed my first VMware vSAN cluster... VMotion'd all the production VM's over from the legacy hosts last night - No complaints, nor did anybody on-site notice

Got the process nailed down, encountered a weird issue with multiple datastores being created - Logged a ticket with VMware, two days later they just said "delete them" ha.

Why would I change from the traditional model of SAN --> FC --> ESXi host to a vSAN?

Am I right in saying the benefits are
+ No SAN admin needed
+ Just present a disk pool and I can integrate it
+Storage is local therefore no need for a SAN

 

[Donny_Dog]

FC is proper 1992.

 

[.Joe]

FC is proper 1992.

FC was great until certain HP switches needed licencing for more than 8 ports 🤔 FCoE took over and the FC can't keep up with the speeds that FCoE can provide now.
HCI is great, need space? just chuck more cheap hardware to tie you over.

with HCI we're squeezing 40 x 3.1G cores 1.2Tb of ram and 38TB Raw SSD space per 1u of rack at sub 10k a pop now
All sat on a 40Gbps back bone

Over the past 10 years gone from 18 Racks -> 3 Racks to 1/3 of a rack whilst increasing space/redundancy.

 

[Gray]

Why would I change from the traditional model of SAN --> FC --> ESXi host to a vSAN?

Am I right in saying the benefits are
+ No SAN admin needed
+ Just present a disk pool and I can integrate it
+Storage is local therefore no need for a SAN

Yup... centralised management, less latency, less physical infrastructure / 'moving parts' in turn means less cost ££ etc. etc.

 

[andybond]

FC was great until certain HP switches needed licencing for more than 8 ports 🤔 FCoE took over and the FC can't keep up with the speeds that FCoE can provide now.
HCI is great, need space? just chuck more cheap hardware to tie you over.

with HCI we're squeezing 40 x 3.1G cores 1.2Tb of ram and 38TB Raw SSD space per 1u of rack at sub 10k a pop now
All sat on a 40Gbps back bone

Over the past 10 years gone from 18 Racks -> 3 Racks to 1/3 of a rack whilst increasing space/redundancy.

It’s not quite as simple as the FCOE is faster than FC.

It’s all about the setup.

I typically run 4*32s together, all active per host.

Also have some mad ninja dedupe and compression that’s guaranteed 3:1 at the storage level

I appreciate that having the data on a back plain is faster but it’s negligible.

It also depends on your use case too. The workloads I typically see don’t saturate the IOPS or have any latency issues. I balance with SDRS for this

I get the benefits of hyperconverged but sometimes a traditional SAN still works.

 

[andybond]

Yup... centralised management, less latency, less physical infrastructure / 'moving parts' in turn means less cost ££ etc. etc.

That’s as I understood it , thanks!

 

[andybond]

Yup... centralised management, less latency, less physical infrastructure / 'moving parts' in turn means less cost ££ etc. etc.

One more.

I assume you have several blades or equivalent packed with disk.

I get that if in a chassis it’s going to be rapid, but if split across sat a series of DL360 then it wouldn’t be quite so quick as has to traverse over a network?

It’s really only suitable for a hyper converged system or blades?

 

[.Joe]

It’s not quite as simple as the FCOE is faster than FC.

It’s all about the setup.

I typically run 4*32s together, all active per host.

Also have some mad ninja dedupe and compression that’s guaranteed 3:1 at the storage level

I appreciate that having the data on a back plain is faster but it’s negligible.

It also depends on your use case too. The workloads I typically see don’t saturate the IOPS or have any latency issues. I balance with SDRS for this

I get the benefits of hyperconverged but sometimes a traditional SAN still works.

Given the budget/opportunity I'd have NetApp/Purestorage or the likes, unfortunately budget limitations push to the cheaper option as it has a much higher ROI off the initial investment.
on a Pure X20 a few years ago i was seeing on average 5:1 and still hitting decent response time.

Management prefer 10k every other quarter to ~250k every 3 years, and i don't plan on arguing management 😁

 

[andybond]

Given the budget/opportunity I'd have NetApp/Purestorage or the likes, unfortunately budget limitations push to the cheaper option as it has a much higher ROI off the initial investment.
on a Pure X20 a few years ago i was seeing on average 5:1 and still hitting decent response time.

Management prefer 10k every other quarter to ~250k every 3 years, and i don't plan on arguing management 😁

I think once you are in , you are in with storage. I know that for £60k (ish) 200TB of Dell EMC storage can be bought and then depreciated over say 5 years.

The snag I can see with vSAN is again licensing. Its another license.

Swings and roundabouts I guess.

 

[Cookie]

Having flashbacks because you've mentioned DEMC storage 😂👀

 

[Donny_Dog]

Flashbacks because no one has mentioned SCSI, or Zip drives 😞

 

[.Joe]

I think once you are in , you are in with storage. I know that for £60k (ish) 200TB of Dell EMC storage can be bought and then depreciated over say 5 years.

The snag I can see with vSAN is again licensing. Its another license.

Swings and roundabouts I guess.

one convenience is the licensing is all under one vendor, thankfully still have a couple of years to reevaluate for the next refresh cycle, the big problem now will be all the big players are either subscription or making a big push to public cloud.

Flashbacks because no one has mentioned SCSI, or Zip drives 😞

reliable tech, just thinking of parallel ports is making my back ache

 

[rctempire]

Anyone else following the vmware fuckshow.

Getting worse now ten percent increase and we don't want you business unless top 500 company.
Disaster for them.

VMware VCSP Customers Seeing 10x or More Cost Increases Under Broadcom

We discuss how some VMware customers are facing 10x or more cost increases under Broadcom with smaller MSPs hit particularly hard

www.servethehome.com

 

[andybond]

Anyone else following the vmware fuckshow.

Getting worse now ten percent increase and we don't want you business unless top 500 company.
Disaster for them.

VMware VCSP Customers Seeing 10x or More Cost Increases Under Broadcom

We discuss how some VMware customers are facing 10x or more cost increases under Broadcom with smaller MSPs hit particularly hard

www.servethehome.com

I cannot understand VMWare rationale behind this.

They made money prior to Broadcom - what are they trying to achieve?