CloudatCost IOWait Issues

Mine was pretty bad last night. I constantly have issues when trying to upload files via sftp or scp anymore (php files usually). At this point I've pretty much given up on CloudatCost. Too bad I can't get a refund for all my Dev and BigDog instances.

Danp

Seems to be pretty snappy this morning. Anyone still seeing I/O issues?

scottalanmiller

Yes, still seeing issues. Here is our list, expanded by one new host added since the last list. Notice that the lowest IOWaits on the list are Digital Ocean (the one DO server that isn't super low is the log server which is BUSY) and the YL server (York Lab). That new server sits on the Drobo B800i iSCSI SAN with RAID 6. So that will give you an idea of how low IOWait should be on a very high latency, low performance, over the network SAN!!!!!

cc-lnx-jump  :  1.23
cc-lnx-ublab  :  3.58
cc-lnx-dev1  :  2.98
cc-lnx-rh7lab  :  1.44
cc-lnx-rh6lab  :  5.51
cc-lnx-mango-st  :  0.33
cc-lnx-dblab1  :  0.52
cc-lnx-dblab2  :  1.95
cc-lnx-dblab3  :  0.79
dny-lnx-log  :  0.19
dny-lnx-pbx1  :  0.02
yl-lnx-elastixmt  :  0.01

gjacobse

Mine are showing a considerable improvement over yesterday. they were in the 20-40 range.


07:20:01 AM     CPU     %user     %nice   %system   %iowait    %steal     %idle
07:30:01 AM     all      2.14      0.00      0.21     18.19      0.00     79.46
07:40:01 AM     all      0.18      0.00      0.02      1.69      0.00     98.12
07:50:01 AM     all      1.96      0.00      0.19      5.94      0.00     91.91
08:00:01 AM     all      0.35      0.00      0.04      2.76      0.00     96.85
08:10:01 AM     all      0.03      0.00      0.01      1.04      0.00     98.92
08:20:02 AM     all      2.39      0.00      0.24     19.76      0.00     77.61
08:30:01 AM     all      0.43      0.00      0.04      1.25      0.00     98.28
08:40:01 AM     all      0.49      0.00      0.05      1.97      0.00     97.49
08:50:01 AM     all      2.56      0.00      0.24      8.81      0.00     88.39
09:00:01 AM     all      0.75      0.00      0.08      7.45      0.00     91.73
09:10:01 AM     all      0.30      0.00      0.03      5.41      0.00     94.25
09:20:01 AM     all      1.65      0.00      0.16      4.93      0.00     93.26
09:30:01 AM     all      1.37      0.00      0.14      3.82      0.00     94.68
09:40:01 AM     all      0.00      0.00      0.00      0.23      0.00     99.77
Average:        all      1.03      0.00      0.10      3.97      0.00     94.90

Please note - right now,.. it's just sitting there. I have nothing running on it,

scottalanmiller

That's a crazy spike!!

gjacobse

I actually have something running now....

08:30:01 AM     CPU     %user     %nice   %system   %iowait    %steal     %idle
08:40:01 AM     all      0.49      0.00      0.05      1.97      0.00     97.49
01:00:01 PM     all      0.70      0.00      0.10      2.85      0.00     96.35
01:10:01 PM     all      6.85      0.00      1.50     31.20      0.00     60.45
01:20:04 PM     all     23.97      0.00      2.22     20.04      0.00     53.77
01:30:04 PM     all     14.11      0.00      3.41     82.45      0.00      0.03
Average:        all      1.61      0.00      0.24      6.44      0.00     91.72

scottalanmiller

OMG, 82%!!!

scottalanmiller

That's literally the highest IOWait I've ever seen.

Dang. I still think These IO issues caused some corruption on my CentOS box somehow.

scottalanmiller

@thecreativeone91 said:

Dang. I still think These IO issues caused some corruption on my CentOS box somehow.

Very possible.