Why I don’t like blade servers

0 Flares Twitter 0 Facebook 0 LinkedIn 0 Email -- 0 Flares ×

After the recent release of VMware VSAN, there has been a series of blog posts from some of my peers talking about the design considerations that VSAN has brought. If you read them in the order they were published, you can follow the conversation that is going on:

VSAN – The Unspoken Truth by Maish Saidel-Keesing

VSAN – The Unspoken Future by Christian Mohn

VSAN – The spoken reality by Duncan Epping

I totally agree with both Christian and Duncan, and in a certain way also with Maish in thinking blade servers are not a good solution for VSAN. My opinion is even more radical, as I think blade servers has almost never been a good solution “at all”… This reminded my of an idea I always had in my mind (and I applied it often in my datacenter designs): I don’t like blade servers. In this post I’m going to explain you why, reason by reason.

Be careful, there is nothing in favor of blade servers listed here.

Space savings? Sometimes

One of the biggest selling value of blade servers has always been the savings in rack space. Compared to a 1 RU (Rack Units) server, a blade chassis can hold more than 1 server per U. For example, a Dell M1000e occupies 10 RU while holding up to 16 servers. In this way, a server uses 0,625 RU. You would say 38% in space saving is a lot, but there is a trick in this number: this value is true only if you load the chassis with at least 11 servers, that is 68% of its capacity. Any number below this makes the rack usage comparable with rack servers, or even worst. For example, the quoted SuperMicro Fat Twin in Christian’s post offers a space saving of 50%, much more than blade servers.

This means, from a design perpective, that with blade servers you need to size your infrastructure with the correct number of servers from the beginning, even if they are supposed to be added “as you go”. At the end, from a scalability perspective, the growth is effective only if it’s made with 10 servers at a time.

Space savings???

Have you always seen completely filled chassis? I did only few times, the vast majority of them are half empty, maybe because the initial requirement was only about few servers and the planned growth never happened, or other times because there are some blade combinations you need to follow inside the chassis, and not all models can be inserted as wanted. In all these scenarios, your blade system is using more RU than the same number of rack servers will do.

Look at this picture a friend published on twitter some weeks ago, do you think 4 servers in 10 RU are saving space?

The chassis is a lock-in

Even if vendors always tell you the chassis is going to be supported for years, and it can accomodate several generations of blade servers, are you sure this is going to happen for real? I’ve seen many customers having to spent so much money at a certain point because they need a new blade model, but the actual chassis was not able to run it.

So, instead of adding only one new blade, they had to add a complete new chassis. And often the price of it is much more than a single blade server.

Shared Backplane

This is by far the biggest complaint I’ve always had in regards to blade servers. I know, I know, modern backplanes are redundant, completely passive, it’s almost impossible they can break. To me, that “almost” is enough to be afraid of them. When a backplane breaks, I loose all of a sudden a huge bunch of servers. If I’m a small company and I only ordered one chassis (like my friend’s company did), I have no other chassis to power up my servers. No matter how many blade servers you have, your single point of failure is no more that single server but the entire chassis.

Datacenter un-friendly elements

When it comes to server rooms design, rack footprint is only one of the elements you need to consider.

First, your servers are not the only component: if your central storage is going to use 4 racks, do you really care about few more RU used by your servers? You can save waaaay more money by optimizing your storage infrastructure than your server infrastructure.

Then, it comes air cooling. Since a blade server has more or less the same internal components of a rack server (mainly the CPU and other internal chipsets), the power consumption is going to be the same, and so it will be the required cooling. But since you are concentrating many servers in a small amount of RU, a blade chassis can become a “hot spot”, and your cooling system needs to take this into account. It’s not a bad thing per se, also SuperMicro fat twin systems have this problem, maybe even more because of internal disks; but you end up designing your datacenter specifically for blade servers. What if you want to install a new chassis in an area of your server room where air conditioning is not enough?

Connections savings

One of the other selling value of blade servers is the savings on connections. You only need to connect to the outside world a single chassis with few cables, thus saving on cabling. This is true, no doubt. But do we still need this? The savings on connections start from the assumption that not all servers are going to fully use the available bandwidth (being it ethernet or fiber channel) at the same time, so you are basically overprovisioning those connections. But with new technologies like Flash Memories, it’s easy to saturate a 10G connection on a single server, so why this would not happen in a shared connection? The solution can be to have a bigger backbone, like 40G or 100G, but in this way, we are still saving money? Or the price of few 100G connections is much more than two 10G connetions per server?

Modern datacenters are embracing ethernet connections for storage too, so the prices of Fiber Channel networks are not a problem, bacause they are simply ignored. And when it comes to ethernet, 10G connections are becoming more and more common as their prices are dropping. Bypassing the blade connections means I have one less component in my data path that can break and one less hop for my data. In some datacenter I saw, VMware clusters for example are spread horizontally, to prevent problems to PDUs. With this kind of design, where also TOR (Top of the Rack) switches are (maybe) less useful, servers are often directly connected to central big switches. And this makes internal connections of a blade chassis less relevant…

What’s the point? As price per network port is falling down, the complexity of some designs is becoming less relevant. In the past, a rack server with 4 gigabit connections was consuming 4 ethernet ports to have 4 GB total bandwidth. New servers with 10G ethernet ports only consume 2 ethernet ports to offer 20GB total bandwidth. As this ports/bandwidth ratio improves by the day, the need for network concentrators like those inside the blade chassis is becoming less clear.

Convergence? Not on a blade

This brings me to my last point. Maish stated correctly that a blade server cannot be used for a converged infrastructure. Not only VSAN, but also other solutions like Nutanix for example are using a totally different form factor, with better space for local disks and cards (and NO shared backplane by the way)

But even before converged systems, a blade servers already had problems in the past to accomodate anything else from CPU or RAM. Any additional PCI card in a blade server often requires a dedicated model (mezzanine or whatever), and you couldn’t simply buy a common PCI card and connect it into a blade. Think about a Flash PCIe card like Fusion-IO or Virident, or a GPU accelerated card. Some would say: you can use the bigger blade models. But then, again, where is the space saving if I buy a “fat” blade server?

In conclusion, I think converged architectures are only exposing even more some of the problems that blade servers has always had.

I know, it’s a radical position; feel free to disagree, but if you work for a vendor of blade servers, please state it before commenting.

28 thoughts on “Why I don’t like blade servers

  1. i completely agree with you: my opinion was unbiased a few years ago, but my boss loves blades (and vendor loves too because of the lock-in you mentioned) so i’m working with them.

    Well, the more i know them the less i like them. For example It’s not as funny to recieve a new blade with a warning on it “before to insert it’s mandatory to upgrade the enclosure firmware”, expecially if the vendor suggests to perform the upgrade “in a mainteniance window”.
    Beware of it expecially if you think to host different companies/customers resources in the same enclosure.

    Finally… a few weeks ago a new enclosure arrived, we took out a blade and… it was not possible to reinsert: a piece of the backplane was broken. To replace the backplane you mus power off everything, pull the enclosure out of the rack, dismount it, replace backplane and viceversa…

    • P.S.: keep in mind that removing a blade enclosure from the rack requires the removal of any component in it (servers, switches, power, fans….); after this three guys (4 suggested) will be able, with a significant effort, to pull down the enclosure.

  2. There is a couple of factual inaccuracies:

    1. Dell can hold 32 blades if you use quarter blades (M420) so you get more density with blades
    2. Most blade makers let you use passthru kits negating the bandwidth issue

    One pro you didn’t make in your post is that racking and wiring up a blade is much faster that a rackmout unit, especially if you don’t buy from a single vendor(which will always have custom rails).

    We use blades extensively and we are very satisfied, we like them because not using local storage(boot via sd) and having a very simple set of requirements(less than two expansion cards) we are simplifying our ops(less time and effort to planned and unplanned mainteinance). There is nothing stopping you from having dedicated rack servers for storage(make it vsan, or starwind or datacore) or vdi(so you can stuff all your grids and firepros) while the low hardware complexity runs on blades

    • Exactly once you get a chassis setup for LAN and SAN you are golden….Need a new server slide it in…boot…configure system…Done.

  3. I would add costs.
    Every time I considered the blade option, costs was higher than the equivalent stand alone servers. Ok, space was not an issue.

    Probably blades were a good option before virtualization, now the reasons you detailed in your article makes them an option not to consider.

    • well said: a blade server without disks quote in the same order of width of a 2U rack server with a significant amount of local, fast storage.

  4. Hey you are entitled to your opinions, and you’ve articulated them fairly well. However, a lot of your points don’t align with what I’ve seen. I don’t work for a vendor or even a reseller. I work for a medium sized company, we run a mix of rackmount and blade servers in a roughly 50:50 split. My views are my own etc.

    “Look at this picture a friend published on twitter some weeks ago, do you think 4 servers in 10 RU are saving space?” Look at the empty space underneath the enclosure. I don’t think this person has rack space issues, so it’s not an issue for them. We have several enclosures in our datacentres (over 20), and almost all are completely full of blades. I know of at least two other companies in the same city who have far more enclosures than we do – all full. You may not have ever seen a full enclosure, but that’s hardly a fair representation of most use cases.

    “I’ve seen many customers having to spent so much money at a certain point because they need a new blade model, but the actual chassis was not able to run it.” Which blade vendor are you talking about? We are running latest gen blades in first gen enclosures and haven’t hit any significant limitations.

    “instead of adding only one new blade, they had to add a complete new chassis. And often the price of it is much more than a single blade server.”. It depends on your vendor relationship I suppose. We have had vendors apply generous discounts to enclosures – in fact we have received several for free. Even if you do end up paying full price, the cost is often offset by the lower cost of blades vs rackmounts, due to power, networking and cooling systems being in the chassis.

    “When a backplane breaks, I loose all of a sudden a huge bunch of servers.” – never had a backplane break in 5+ years of running blades from different vendors. We’ve had other failures, sure – but they’ve been limited to a single blade. I think your fears are unfounded. Modern enclosures don’t have single points of failure.

    “What if you want to install a new chassis in an area of your server room where air conditioning is not enough?” – I suggest you don’t do this. All datacentres are not created equal, you should pick your hosting provider carefully. If temperature control is an issue, bring it up with them, don’t blame your server density.

    “As this ports/bandwidth ratio improves by the day, the need for network concentrators like those inside the blade chassis is becoming less clear.” – it depends on your use case I guess. It’s far nicer to have two 10GbE cables per 10U chassis / 16 blades than it is to have the equivalent for rackmounts.
    Also a server admin can move a blade to a different VLAN without having to get the network team involved, or worse, a site visit.

    You missed one of the benefits of running blades over rackmounts – ease of management. I can log on to an enclosure and get inventory of all the servers, look at hardware issues, even get remote consoles onto them if I need to. Most vendors let you join enclosures together, so you can log onto a single enclosure and get a consolidated view of a whole rack.
    There’s no easy way of doing the same thing with rackmounts, is there?

    • Mark, I completely agree with you – especially on the ease of mgmt. part. in regards to deployment, inventory, troubleshooting etc.

      Blades are not all bad. 🙂

    • I was just reading through the article and came to Mark’s post and agree pretty much all the way with him. I think for small businesses, the context in which the author might have developed his perspective, blades aren’t likely going to be the right choice.

      I’ve worked in Government departments where they typically have two data centers in their home city and upward of thirty racks in each data center. The primary vendor was HP in this department and each rack had three enclosures. Very few of them were empty.

      We only used rack mount servers for larger applications, like databases. One of those servers would have 2+ TB RAM and 10 sockets (total 80 cores). We were building a blade-only section when I left (about two years ago) using an intelligent cooling system that was cooled by cold outside air temperatures for at least six months of the year, giving us great power savings. This was combined with a locally made rigid cold-isle containment system.

      We also moved to larger racks that could contain two top-of-rack switches and four HP BladeSystem c7000 enclosures. This would give us sixty-four half-height dual socket blades per rack! Each rack was setup in an identical manner and once in place the infrastructure didn’t change for years.

      Because everything is so nicely contained we also did not order front doors with the racks and we could make the isles a bit narrower. This is great when data center floor space is expensive.

      I really like HP blades, and I am keen to work with Cisco’s blade infrastructure, too. However, they are good for medium sized infrastructures and smaller and larger would probably benefit from other server types. Actually, for smaller requirements, putting everything in a cloud service, like Azure or Amazon, would make much more sense to me.

      With Intel’s 3D XPoint storage technology becoming available sometime in 2016 (fingers crossed) I think we will see a significant change in our data center and HCI vendors like Nutanix beginning to look even more attractive to mid-sized infrastructures that have traditionally been biased towards BladeSystems.

      One thing I really don’t like about a BladeSystem architecture is that there is extremely limited choice for network hardware and design. Ideally we want to keep networks in the data center very simple and very fast, and virtualise on top of that. Cisco aren’t interested in innovating their HP BladeSystem network products and who in their right mind would use HP’s network products.

      The problem with Cisco’s Nexus product is that they want to sell you more than you really need, which means more than two tiers in your data center network design. If you look at the Nexus products you’ll find yourself going around in circles trying to build that fast, simple network, and realise that they’ve made it so you can’t.

      None of the vendors want you to be able to build with less by taking advantage of faster CPU and network bandwidth. So, they’ll build convoluted SDN technology into the hardware and leverage that to develop new licenses. None of them want us to run lean.

      I’d like to see Nutanix do a hybrid BladeSystem or just design a rack-level backplane for power and communications to enable more efficient rack usage. There’s gottabe some juicy innovation to be had by taking the best of both worlds. Yeah, the rack is due for a redesign, too.

      • Thanks for sharing your experience!
        In regards to the last point, you may be interested in looking at some research done by Intel about their Rack Scale Architecture, it’s similar to what you are asking for.

        • Thanks for the tip on Intel’s Rack Scale Architectur, Luca. It looks like a fairly comprehensive suite of technologies. I haven’t seen the physical rack redesign yet (assuming that they have one). Very curious.

  5. I completely agree with your points! I’d like to add some additional points and my experiences with blade servers.
    I think blade servers are not the ideal form factor especially when used for virtualization purposes.You are very limited when it comes to PCIe expansion/HDD slots, you introduce additional complexity when it comes to your HA/DRS cluster design (anti/affinity rules) and you are sharing the uplink bandwidth on each chassis to your core and SAN switches again if you want to save some ports. My previous employer also had a policy to only fill up each blade chassis half-full so we could quickly move blade servers to another chassis if one would fail. Therefore, the space savings argument wasn’t valid anymore. The costs of this design restriction vs. rackmount servers were therefore much higher. We also had to change the datacenter to a hot/cold aisle layout to remove hotspots of the blade chassis. At that time (2009/2010), HP also removed a quad-port ethernet card option on the successor model (BL460c series). So we had two options: go for 10G (network wasn’t ready at that time) or change the blade servers to a different model which would fill up the whole slot. I dont want to blame blade servers, but I think you are much more limited than with normal rackmount servers.
    I think if you want to be more flexible and are not constrained when it comes to rack space, rackmount servers or these new hyperconverged/scalable systems like the supermicro twin or HP SL2500 are the way to go.

  6. Hi, I’m Claudio and IWork4Dell.
    I’ve never been a huge fanboy of blade servers, mainly because they add a complexity layer most people could do without. Therefore from that point of view I share your sentiment about those devilish contraptions, but I’m afraid this post isn’t a hundred percent fair.

    1. Just for starters, space savings isn’t a trick. People buying a 10u chassis for 4 half-height servers isn’t the servers’ fault. Nobody prevents me from renting a thousand sq. meters datacenter for a single R420, but then I can’t blame the poor server.
    It CAN be sensible in case you’re going to add one more each week, for six months. Or for full height systems.

    2. The Poweredge M1000e supports now three servers’ generations: 10th, 11th and 12th, and features keep being added and/or expanded all the time. There’s no talk whatsoever about dropping support for many more years, and considering it came out on jan. 2008, it’s 6 years support since (and counting).
    If a vendor fails to deliver, perhaps it’s time to choose another?

    3. I’ve supported enterprise hardware and software for several years and I MAY have seen a midplane replacement or two over the years. I say may because I frankly can’t remember, but it must have happened once or twice.
    When a midplane breaks, it doesn’t disintegrate. Perhaps a port goes down, and you may lose communication over one port, over one fabric.
    Alright, let’s be pessimistic. Let’s say it’s TWO blades, even TWO fabrics. Everything else keeps going on like a trooper.
    The only downside now is having to schedule a 20 minutes downtime for the part replacement.
    Right there, if you “only” need a 99.999999% uptime, you can’t afford to have a single chassis. If anything, you _need_ a _very_ robust business continuity plan (which possibly means a spare server or two, if not several more chassis).

    4. “The savings on connections start from the assumption that not all servers are going to fully use the available bandwidth (being it ethernet or fiber channel) at the same time, so you are basically overprovisioning those connections.” Right. And how is that different from consolidating ten servers in one, using virtualization technologies?
    Now, about costs. I honestly don’t know what to say. On one hand I can read that blade servers don’t offer enough connectivity flexibility, and the solution would be a bigger backbone, but that would increase costs.
    On the other hand, though, I can read that now the cost per price has dropped significantly.
    I’m confused.

    5. “Maish stated correctly that a blade server cannot be used for a converged infrastructure.”.
    I can confirm we did set up several “proof of concept” on converged networking in order to demonstrate the capabilities of an S5000, for instance, and we always used blade servers just for the convenience.
    For use cases with solid state storage cards or GPGPU, well…it’s simply not their playing field nor even the same sport. Some Poweredge C5220 with C410x may be a better alternative.

    To sum up: yes, I work for Dell, a widely known vendor of blade servers (and more), therefore I may be (candidly) biased, but my impression is that I’m not the only one.
    Not all datacenters pivot around VMware and VSAN. Some have different needs.
    A famous quote, widely misattributed to Albert Einstain, goes: “Everybody is a genius. But if you judge a fish by its ability to climb a tree, it will live its whole life believing that it is stupid.”
    The truth is that [blade servers | virtualization | FcOE | AnyTechnologyOfYourChoice ] is great, but only if it’s used where it’s useful.


  7. Hi,

    What I like in blade servers is integrated management – but some rack servers also offer this option (and, of course, additional tools can be added to manage any servers).

    Backplane replacement is not very likely by it can happen – I worked for a year in company having 100+ blade chassis – and I remember 1 (one) case when backplane had to be replaced.

    I also remember other case when power outage caused both chassis management modules to fail – servers survived, but modules needed to be replaced – so there was outage for whole chassis.

    On other hand there were significant numbers of issues without clear root cause – when blade reseat helped.

    So in my opinion, blades are useful when large numbers of servers are installed – having only one chassis can be bit risky. Having two – to be able to recover from failure only half of space can be occupied in each. So I think one should start considering blades when there are servers to fill in 2 chassis (and then buy 3).

  8. Thanks everyone for passionate comments, I knew from the beginning this was going to be a hot topic, so again thanks everyone for jumping in and explaining your ideas.
    I already said what I think about in my post, so no reason to comment back, only one small clarification to Dell guys: I picked up Dell as an example because my friend sent me that picture with a Dell chassis. It’s not against Dell, I could have chosen any other blade vendor, my opinions would have been the same.

    • So you don’t have any actual experience of using blade servers in a datacentre environment?

      • I have no idea how you can come up with this assumption, I only wrote I used that Dell picture because it was the reason that made me wrote this post. Yes I used many blade systems, but when possible I’ve always stayed away from them in my designs, for the reasons I explained at length in my post.

        • It’s swings and roundabouts. For some implementations, blades fill a great technical gap, but stating that you will or won’t use a particular technology as a rule of thumb is, in my opinion, shortsighted.

          I have installed probably 200-300 blade chassis in the last ten years, along with 50-60 Sun large-scale servers, and a thousand or so discrete rack-mount servers. It’s horses for courses.

          You use the technology that best fits the design brief, whether that is for grunt, connections or redundancy.

          I have designs where blades didn’t fit, but when I am looking at a situation where high numbers of, for example, remote desktop sessions are required, blades do a great job. High-powered blades with large amounts of RAM and using a fibre-channel/SAN infrastructure running a virtualisation environment are, similarly, great value for money and great from a manageability perspective.

          My assumption was based on the fact that there are so many potential solutions to a hosting problem, that to do away with one so arbitrarily seemed to be coming from a position of unfamiliarity.

  9. Fully agree with Luca
    Working in Italy I can assert that a little 1 to 3 nodes VMware Infrastructure it’s enough for the SMB customer (90% of Italian business I think). When I see a blade I blame the sales man that sold it, because generally is empty and similar to a black hole (no one at the customer site generally know how to configure networking).
    One plus of Dell M1000e is the ability to embed two Equallogic storage inside the chassis. This is nice, but due to the Equallogic iSCSI requirements (LAG between switches) the network setup of the chassis is complex when you have just to switches on the chassis (typical sell)
    Another plus of Dell chassis is the ability to soft-partition 10 Gbit ethernet interfaces between LAN, iSCSI and vMotion traffic so each service can have full bandwidth when needed (best effort). On HP chassis I’ve seen that the partitioning is hard so I have a customer that need 30 to 45 minutes to put in maintenance mode a single VMware blade with 192 GB RAM because the vMotion pipe is partitioned to 2 Gbit out of 10 Gbit. Hope things have improved on HP sides with latest models 🙂


  10. All of you have valid points… Blade servers have their value and their shortcomings. This is why most Racks will have a mix of servers. Density and power savings are becoming the driving force, but there are times when your infrastructure needs a boat load of PCI slots…

    In all reality, the whole compute world is changing to “core density.” The Nutanix and Simplivity guys are recreating a market that had existed in the Blade world. By bringing storage (high iops), and dense core counts in more reasonable incremental costs… 2RU at a time. HP however IMHO, still reigns with their Moonshot box, and more affordably the SL2500 up to 96 cores and 400k iops in a 2U. That is compute density…

    I think that the 2U Virtualized game is where all of this is going. When you can pool resources and drive down incremental cost, DC folks will start buying in. Any SimpliVity or Nutanix experienced folks like to chime in??

  11. So because vsan is here and almost all of the data centers are now going to use it for their applications – they should ditch blades or any form factor which isn’t suited to vsan?

    Good article and good points – but as someone said amongst the comments – technology is good when it is used appropriately

  12. Fully agree with you Luca. I am not working for any vendors or re-sellers. I work for a mid sized educational institution.

    I have one more point to add. As the complexity increases, fixing up issues will take longer time. I had a recent incident where two servers on a Hp Blade enclosure was behaving odd. Had a case with HP support. After initial log analysis, the first fix they recommended was that servers need a firmware upgrade.

    I did it – Issue not fixed. Further troubleshooting pin pointed that this issue may be due to a faulty LOM. Replaced LOM as they recommended – Issue not fixed.

    Since we had two Flex fabric modules, I suggested the onsite engineer to try out by keeping one Flex down. As all servers are having team and connected through both the Flex, We confidently did that test. After few testing, We concluded that the issue happens only while traffic is going through the second Flex module. HP support gave a replacement. Before the replacement, the engineer suggested that we could try re-inserting the second Flex module. We tried that and the issue is no more.

    What ever be the reason, to identify that the issue is caused by Flex module took a week. And I am sure that if we confidently did the testing with the help of that Onsite engineer, the support would take even few more weeks to pinpoint that.

    • “What ever be the reason, to identify that the issue is caused by Flex module took a week. ”

      I would have fired your incompetent ass.

  13. Oh dear, where to start….

    Blades filled an important niche a few years back. They were a valid substitute for 1 and 2 u servers in the reasonably modern data center. Mount a few chassis, wire them up well and you can plug in what you need. I’ve seen this work well with mixed environments and even with x86 and IBM power blades.
    No, you couldn’t maxx out a network connection and yes they were a bitch to cool or even power in a poorly designed DC. On the other hand, the average server in 2008 didn’t even use much more than 30% of a 100mb network connection and poor power would also prevent you from adding more than 30 servers to a rack.

    Fast forward to today…

    The average datacenter now is rapidly losing servers. We’re moving towards 2 and even 4u servers which are maxxed out on CPU and memory and run ESXi. There’s minimal disk space on the server itself if it’s still there at all.

    Blades are still there for those few things that don’t run virtually because they’d cause hotspots in the landscape or because of licensing issues.

  14. I completely disagree. I have a bunch of server racks populated with a bunch of HP blade servers, all of them running virtual servers. I couldn’t get anywhere near the density of VPS using conventional rack servers for the same price.

    And managing them using HP’s great blade management toolset is a breeze.

    Blades for me every time.

    • Eugene, I’d be interested to know how you’ve achieved some savings against regular rack-mount servers. I’m wondering if all that regular cat6 cabling used for each rack mount server doesn’t add up to significant power consumption vs low power direct-attach SFP+ DAC cables used in BladeSystems, and the power saving capabilities inherent within a BladeSystem, better cooling efficiencies, etc? Care to elaborate on some of the stand-out advantages?

  15. my husband is a very rich and welding man.will make the money together few month later,he started hooking up with bad friends .on my noted he was having an affair with another woman .the family lawyer call me an asked me if me and my husband had a miss-understanding ,because my husband has change the name writing on the wile.he took
    everything we have to the his girlfriend ,meaning that i don’t have any share in the family.i was frustrated and discourage.until a friend of my advice me to visit a spell caster so that all my problems will been solve within 48 hours then i contacted the spell caster she introduce to me.dr ogun spellcaster,drogun promise that every thing will been alright.few weeks later my husband came back home ,on his kneels begging,asking me forgive and forget about the past and face the future ahead.right now i am in full control of my husband access.a big thants to dr ogun who bring back my husband .if you have same problem kindly contact dr. ogun in his via email.drogunspellcaster@gmail.com.

Comments are closed.