Docker For Mac Slow Sync

Docker For Mac Slow Sync Rating: 4,5/5 4625 reviews

Feb 16, 2017 - Docker for Mac and full sync on flush issue. It's so slow that you can even get composer timeouts on “composer install” or “composer update”.

Posted by1 year ago
Archived

Make Docker on Mac faster

I am using docker for all my development. I used to have a ubuntu laptop for my primary development machine, now I switched to Mac. Performance is much slower on Mac though. For an app i managed to bind the volume as read-only as the PHP app is sandboxed and does not write to disk. But instead if I work with Laravel or other stuff in a container, volume speeds are low.

I am using NFS with d4m-nfs and the speed improves a bit.

Also I downloaded the edge version for D4M and tried to bind the volume as delegated. The speed improves a bit more but only when doing a command for two times. The first time is still slow.

Here is a short gif demonstrating times running a simple php command first time and second time. Second time is in 0.4 seconds, faster, but still a bit slow (0.01 on host).

6.0.4 platform packages • • • • The binaries are released under the terms of the GPL version 2. Version 5.2 will remain supported until July 2020. Mac os image for virtualbox. Please also use version 5.2 if you still need support for 32-bit hosts, as this has been discontinued in 6.0.

Steam account for mac. MacOS on Steam Browse the newest, top selling and discounted macOS supported games New and Trending Top Selling What's Being Played.

Are there any things i should do to improve the speed on Mac? And please, don't tell me to switch to linux :). I am ok with mac, just want a little more juice form Docker on Mac.

82% Upvoted

I understand that this is a new feature, but what about people that want to support development on multiple OSes. RHEL 6 and 7 are going to be around for a long time still, so you're saying people will need to keep multiple configs around for years just so:cached can be specified on their Macs?

Well, yes and no. Docker is despite all the hype pushing for it's use, it is currently not mature enough for early majority users to use it broadly yet, has some known stability issues and a fragmented market. RHEL 6 is even late majority users, they by nature don't want to test out new things, so don't make them. RHEL 7 however is supported by, so once they migrate there they can take advantage of this. In other words research usage of docker for future version of your own software, not for older versions.

At least not under all possible supported platforms you have there, that won't blend. On Fri, Apr 28, 2017 at 6:03 AM, carn1x.@. wrote: Is there any high level estimate for when:delegate might start to appear? Just wondering if I should try and get all my devs converted to something like docker-sync which has a lot more initial/ongoing overhead, especially for our devs who are just putting up with our decision to enforce docker rather than enthusiastically embracing it! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

For sure it would not, and not even because this implicates, that the container has the superior over the host - which is right the opposite what you need when you actually develop. The container will hardly ever surprise you with something 'new interesting' - and if anything, its generated and reproducible. Very very different for code changes, which are not reproducible and very hard to write again ( mentally alone ). Thats why docker-sync uses always 'hosts wins' while still using the buffer. Losing code is just not funny - losing generated assets (in case it ever happens to be a conflict) is a no care. So:delegate is just something you do never want to touch - even if i do get the general idea. Subscribers to this issue might be interested to read the following blog post: which explains how and when to use cached, along with some details about the implementation.

Since there are now significant improvements in the use cases that prompted this issue, we're closing the issue to further conversation. We'll continue to post occasional updates here when further performance improvements are available. There's a new, more focused issue open to track the progress of the caching implementation (i.e. The implementation of delegated, and further improvements to cached): If you'd like to follow developments closely, we recommend subscribing there. Thanks to everyone for contributing to this conversation!