Over the last few years, I’ve been doing charitable work upgrading the computer systems and network at a church run hospital in the eastern province of Zambia. This work has been supported by the Zambia Anglican Council Charitable Trust (ZACCT) in the UK, the Medical Support Group (MSG) in the Netherlands and private individuals to whom I am truly grateful. My own company, Realtime Trains, and others we do work for and supply us have also supported the project with funding.
Last August and September, I assembled a team of people from various walks of technology to aid a major week-long project to fully upgrade the infrastructure and replace a series of daisy-chained switches with poor quality equipment with a brand new system intended to last many years to come. The entire team came at their own cost or were sponsored by their employers to help.
The work involved replacing a lot of old consumer-grade equipment with new, or second-hand and refurbished, professional grade equipment and laying brand new cable around the grounds to ensure a robust network that works for the hospital.
Prior to my initial work in August 2015 (I first visited in 2014), the hospital was wired up with poor quality cable sourced locally with Linksys WRT54GL routers and the odd Netgear or TP-Link switch here and there. In 2015, unaware of many of the cable issues (causing massive broadcast storms and associated congestion) all over the place, I set out to replace the core equipment replacing the routers with a full firewall and add a proper server, donated by Bytemark, to replace an ancient one sitting holding medical records and accounts detail. We installed a small UPS to keep this running during the then few and infrequent power cuts, around once or twice a week for short periods.
The work in 2015 also set out to replace a number of small managed switches to try and divide the network up somewhat to try and ensure that there was a level of separation between various uses of the network. On the hospital’s side, there was a broad lack of awareness of what the network actually entailed – and how many nodes existed(!) – so this work wasn’t able to be completed in full due to a lack of equipment.
Since all the work in August 2015, the power situation in Zambia has worsened owing to poor rains in 2014/15 – 95%+ of its power originates from hydroelectric dams – and daily load shedding (power cut) of up to 12 hours became the norm rather than the exception. This, together with being armed of knowledge about the situation in the hospital as at August 2015, led to the major project in 2016.
This project replaced all poor cable and created a ‘fibre ring’ of connectivity around the grounds to ensure that if there was a break in one area we shouldn’t lose anything. This proved to be rather useful in December, but I’ll touch on that in a bit. We also installed brand new outdoor grade copper cable across the grounds to ensure that everywhere was linked in properly. In various buildings we also fully wired up the rooms for ethernet and wifi connectivity, together with configuring fully managed switches across the site. This can all be administrated remotely while staff on site are trained up to work on it themselves.
The hospital sees a lot of doctors from Europe staying for periods of at least 6-12 months – sometimes up to a few years – to gain experience and supplement the staff on the ground. I felt that it was necessary to not only support the hospital, but also them, and so we installed wifi in the mess where they are able to take their meals, relax and revise. All the doctors are housed on-site by the hospital nearby to the Mess so this was intended to aid them as, normally, they would be using 3G dongles (mobile networks are really, really good there, but not cheap!). The hospital also welcomes medical students for short placements.
All of this work was carried out in conjunction with a new satellite based internet connection being installed at the hospital, at approximately 10x the speed of the old one, with caching and filtering being installed at the central core to make it a bit more stable for everyone. We installed new core switching across the site, another new firewall and a low power server. We also installed new UPS equipment to ensure that the systems would not go down in the event of any power cut of less than eight hours to ensure that people could continue working on laptops over wifi.
I went out to the hospital on for a brief scoping mission in May prior to the work to ensure we had the correct quantities of items and to ensure that preparatory works prior to our arrival were made. This work included building cabinets and digging trenches for us to pass cable and conduit, as well as obtaining said conduit, to speed up the process for us when we arrive. Unfortunately, the preparatory works were just being started when we arrived! We had budgeted in the region of laying one fibre cable per day but the delays set us back and we ended up running cables the entire time we were on site.
The delays meant we couldn’t complete everything that we had set out to do, including replacing some computers on site with those donated by a school in Teignmouth. The computers are still on site and we are working on a way of deploying these so that this part of the works can be completed. We also didn’t complete configuration of another donated server from Bytemark but this was installed and we can remotely manage this with the new connection in place.
In late November, we noticed from our remote monitoring that we had lost one of the main fibre links between two portions of the site, as well as a copper link to the building on the right. A tree had collapsed overnight, apparently rotten and hollow, onto the runway causing significant damage to the roof and had severed the cable entirely. Fortunately, our loop ensured that connectivity, barring to that one building, and another linked from it, was completely intact after a few seconds of reconfiguration.
We have subsequently had a few issues and lost managed connectivity to the administration block – our second core – owing to a configuration issue (we are not sure where this originated as the switch in question has not been touched since installation) and are working on solutions to resolve this. It unfortunately looks like this may require new equipment.
With the replacement of a fibre cable, and more copper cable being required, I am currently preparing a shipment of equipment over there and would appreciate any money that you may be able to contribute towards the equipment. Any additional funds will go straight to ZACCT for other projects they do with the hospital, such as supporting local AIDS orphans or towards medical supplies. We may also be looking for a small team to go out again in 2017 or 2018 and perform remedial and extra works – if you are interested, please contact me.