Planned downtime: Server move (2019-02-27 13:00 UTC)

[MagicManICT]

After move, why I feel more lag?
I have ping around 100, but still always repeating one action for several times then jump to another place where I click before...

There's another issue in server processing loftar is trying to track down. It may be a bad piece of hardware, but he wants to verify before committing to it and it doesn't work.

[loftar]

If it's a Intel SSD, it is common with them, they have a cache problem, if you can choose at the datacenter what SSD to pick, go for SAMSUNG

It is not. They seem to be Toshiba drives, and no, I don't think I can choose.

better buy your own for a good brand like DELL or HP, and host it in the most convenient datacenter, in the long run, it's cheaper, and you have more control over your hardware.

I don't disagree in principle and I would really like to do that, but it seems to be really difficult to find reasonable colocation around here.

[KeTo]

seems like freeze'o'lags just became stronger and longer =\

[Hollogodis]

[dullah]

- Always use a good RAID adapter, PERC H730P or 740P on modern DELL servers for example, do not go with RAID software.

I disagree with that. Use a good HBA and a storage stack (eg. ZFS) that checksums all the way down so it knows what part of the data is actually good.

I tend to disagree, performance-wise the HW Raid is unmatched for intensive IO workloads.

[Granger]

- Always use a good RAID adapter, PERC H730P or 740P on modern DELL servers for example, do not go with RAID software.

I disagree with that. Use a good HBA and a storage stack (eg. ZFS) that checksums all the way down so it knows what part of the data is actually good.

I tend to disagree, performance-wise the HW Raid is unmatched for intensive IO workloads.

He who sacrifices data integretiy for performance deserves neither.