What is your SLA? Can you guarantee a 99.9%? A 99.99%?
PDFShift has been around for more than 7 years now, and we've seen our fair share of outage.
We had to handle our data center catching fire, database going down, bad deployment, and things that were totally unexpected.
We always took these downtime as a blessing in disguise; we see our service as critical, and something that put us down should never happen again. So any downtime was an opportunity for us to ensure this will never happened, and every downtime we faced increased the quality, resilience of PDFShift.
That's why we now have redundant servers on different cities, multiple databases replications with delayed writes, unit tests, and a ton of fine tuning to handle all these drops in performances that we fought during those years.
Our status page currently shows a near perfect 100%. We have some servers that had a hiccup, but since they are redundant, the whole PDFShift service hasn't been down for quite a long time now.
But despite all this, we can't, and we won't guarantee any level of SLA, being at 99.999%, 99.99% or event 99.9%, and the reason is simple;
Despite being one of the most important company in the world, Cloudflare recently had massive downtimes that were absolutely unfortunate and hard to solve.
If a team of the best engineers in the world can have hours of downtime, how can we - a definitely smaller team than Cloudflare - can guarantee that no, we will never have a downtime of more than an hour ?
That would be lying.
And we don't want to lie to you.
What we can promise you, is that when downtime will happen at PDFShift, we will be proactive, jumping on the issue as soon as possible, and pour our heart out to solve this in the least time possible so that your service will be impacted as low as we can.
That's something we guarantee.