Hopefully this helps someone out there, and at the very least points out that there are still serious issues with 2019. Of course the strange thing is that a 2019 VM ran even slower than a 2008r2/2016 VM on a 2019 host, which I would think would eliminate most of the driver issues on the VM level. Now if someone has any ideas of what's wrong or changed in 2019, I'd love to save having to rebuild the entire cluster. Speed Up CrashPlan Backups And Free Up CPU Power With These Scripts. So with that, I scrapped 2019 completely for 2016 on the host, and sure enough it ran 74 / 65 right out of the gate. Finally it occurred to me to fire up an old 2008r2 VM on my 2019 hyper-v host, and to my surprise the number were a good bit better at 25 / 30. Based on the limited speed that CrashPlan allows uploads, it would take over a year to add the missing 5 TB. Anything above would be deleted and needs to be backed up again. The Q1T1 for example would run 21mb read / 24mb write on the new server, and 35 / 38 on the old server. But the website says you can only migrate up to 5 TB to the new plan. The fastest way to test and compare the problem was to use crystaldiskmark, and look at the 4k random tests, since sequential tests performed fine. A folder with thousands of small files would copy over on local disks in 23 seconds on my old server, but take 2 min 40 sec on the new servers. The problem was mainly with small io intensive tasks. I then spent the next three weeks tweaking every windows and bios settings I could find, updating everything, and purchasing different Nics, raid controllers, and drives in an attempt to fix it. Using crashplan as a secondary backup, so speed was not really a concern, but with several TB of data, the time it was taking was ridiculous. I noticed right away that these servers were running slower than our old servers, with poor mysql performance and folder transfers. just tried this fix on Server 2008 R2, speed went from 2.5 Mbps to sometimes 84 Mbps (depending on file). I would certainly setup a Timemachine backup first, day 1, before initiating the Backblaze (or other) cloud options.I recently built a cluster of new to me Dell R620s, which run HyperV with various services. If you want to be able to go back and rescue deleted files months after they were originally deleted, then TimeMachine and Crashplan are better options. Other options like Crashplan Small Business keep deleted files indefinitely. Even under ideal conditions, the initial upload process can take a while. Designed to make minimal demands on your computer's resources, the CrashPlan app default settings favor quiet. The problem is the first FULL backup of course which may take 2 to 5 months.īackblaze does have a facility to maintain deleted files for a period of time (from memory, I recall it being 30 days), after which the deleted files are gone forever. Speed up your backup using CrashPlan Overview. Review the volume of data being backed up for first time backups or the volume of the update for incremental backups. The backups are similar to timemachine in that, once the first FULL backup is complete, it only backs up "net changes" on a daily (hourly?) basis. Are there any work arounds for this problem? Also, is the backup process similar to Time Machine: only new data is backed up, anything deleted on the hard drive since the last backup disappears in the new backup? My upload speed is maxed out at 10-15 Mbps. I'd like to use online backup, like Backblaze, but I have 6 TB of photos and I might die before getting it all uploaded.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |