We have been using Moodle for a couple of years, and have it installed on our server.
We have asked our IT providers (a third-party company) to upgrade the application to the latest version of Moodle.
They have responded that it will take 25-30 hours.
We have not made any modifications to the standard application in the previous not upgraded version.
Does this sound reasonable?
Regards
It entirely depends on the original version as to how hard it will be to upgrade.
25-30 hours, however, seems to be excessive, the longest time is generally taking pre-upgrade backups, the actual upgrade process it's self is very fast, a matter of minutes.
Depending on the backup size etc, that may take an hour or more, but considering we upgrade sites like moodle.org in an hour or so, I don't see too much time being used doing an upgrade.
You could always take a copy of your current instance, and upgrade that first to ensure everything works and also get a proper time estimate for upgrade (we also do this before production rollout).
Upgrading is very simple, and the process is documented well on our docs.
Alternatively, you can seek a quote from one of our many partners whom would be more than happy to help.
Hope that helps
Matt
This depends on a lot of factors, but this estimate does seem high. It's hard to say what processes other companies may do but here are my thoughts.
I would estimate for one of our hosted customers there may be 1-2 hours work involved in a minor 2.x upgrade. This could increased slightly if I'm working outside of our own hosting environment, or if I have to backup data over a slow connection (although you may expect to only be charged for human effort and not the time taken for data transfers to occur).
To migrate and upgrade a new client from their old hosting environment to ours e.g. a 1.9 site all the way to 2.9 I would allow between half a day and a day, so 4-8 hours. If your chosen IT providers don't have a scooby-do about Moodle you can extend these time scales to account for them fumbling and making errors.
You should ask them to break up the estimate into specific operations, e.g.
This said, some HE clients we support may easily take 30 hours of work for their annual upgrades, involving UAT environments, multiple third party plugins, local modifications, archive snapshots, the coordination of several departments and services at the institution and 100's of GB of data being moved around. Most of the time is planning and testing and the actual 'upgrade' typically takes a morning.
Where I work this is all rather moot though, as we don't charge for upgrades as we see them as a necessary part of sustainable Moodle hosting
Although you might not have made modifications, you might plugins, and setup authentication to external systems etc.
Sometimes the only way to ensure that the site is going to operate after an upgrade is to make a full copy and test it first - which double the effort required.
You don't say which versions of Moodle - so you could be on 2.2, which would require multiple upgrades to get to 2.9.
There are a lot of factors, but I agree that 25-30 hours sounds excessive (unless restoring courses manually into a new environment) - and I'd expect something ranging from 5-15 hours as a more 'normal' figure. However, as stated above, there are a lot of factors that can affect the time required.
I agree that that is excessive. If you version is two years old, you are probably running 2.6 or 2.7. If 2.7, I can't honestly see more than a couple of hours, maybe three if running 2.6. The longest part is running backups with I imagine you have full site backups anyway...
"I imagine you have full site backups anyway..."
That is what everyone thinks until just after the disaster
The amount of time sounds more than I would expect, but not necessarily by much. When doing an upgrade I would do a practice run on an entirely different machine then some time would be taken with testing. I don't mean extensive end user style testing but an hour or two "sanity checking". In my view everyone underestimates how long things take and if you do that when you are charging other people by the hour can easily go out of business.
When I first saw this post I assumed there had been reasonably extensive plugin customization, now that can be a massive time sink.
I don't think anybody can answer this unless you provide a breakdown of what they intend to do with their 25-30 hours. If you don't know, then you need to find out.
On the face of it it does seem a lot but we don't know what they intend to do. The (careful) preparation for some very large/critical sites I deal with can take considerably more than that!
