49,519 reputation
28161324
bio website kbrandt.com
location Hollywood, FL
age 32
visits member for 5 years, 7 months
seen 3 hours ago

I'm the director of site reliability at Stack Exchange. I'm also the co-author of the bosun monitoring system and sometimes blogger on the Server Fault Blog.

@KyleMBrandt


Nov
30
comment Choosing the right AWG for Server Power Cables?
Doesn't voltage matter as well?
Nov
30
comment Choosing the right AWG for Server Power Cables?
@MDMarra: So as not to be confused with the floor units. From wikipedia, "The term (PDU) may refer to two major classes of hardware power devices; the first and typically the general unqualified term refers to the category of relatively higher-cost floor-mounted power distribution devices which transform one or more larger capacity raw power feeds into any number of lower capacity distributed power feeds .... The second type of PDU is sometimes called a Smart-PDU, Rack-based PDU, Intelligent PDU or simply "Power Strip" by various IT professionals"
Nov
29
comment How do SAN Certs degrade performance?
I'm testing some self-signed SANs now, can't get a measurable difference with HAProxy as the server and CURL/WGET as the clients with 2.5k names. Tried 20k names but got (35) error:1408E098:SSL routines:SSL3_GET_MESSAGE:excessive message size
Nov
29
comment How do SAN Certs degrade performance?
On a call with Comodo they told us this was the case (starts to degrade over a hundred or so). I was guessing it was maybe "processing each entry", but I'm not sure of the processing flow there -- hence my question.
Nov
28
comment What does Active, Targetset, and Active targetset mean in the Output of dfsutil /pkiinfo?
@tonyroth: "OR" The client is in a different network (subnet) within the site than OR-UTIL01, but the client's subnet is properly assigned to the same site in AD Sites and Services.
Nov
28
comment What does Active, Targetset, and Active targetset mean in the Output of dfsutil /pkiinfo?
I did read "NOTE: PKTINFO indicates a successful connection to the referral returned by listing it as ACTIVE" So maybe NY-FS01 is blocked by the firewall (pretty sure it is from this client). Just Active means the connection is working, ACTIVE TARGETSET means it is Active and that connection is being used? But what would just TARGETSET mean?
Nov
14
comment Instances in OpenStack are not getting DHCP leases
@LorinHochstein: Seems to exist already at bugs.launchpad.net/nova/+bug/1029430 , someone was asking about this in #openstack too, which was why I posted this. See (github.com/mseknibilel/OpenStack-Folsom-Install-guide/issues/14) for more info on my setup.
Nov
14
comment Instances in OpenStack are not getting DHCP leases
@LorinHochstein: I wasn't getting leases with the Cirros test image or the latest Ubuntu images, so I don't think this is limited to only a obscure guests. I do have libvirt_use_virtio_for_bridges=True so I don't think that would fix it.
Nov
2
comment Autoaccept all Certs on Puppet Master
This is on the master?
Oct
20
comment HAProxy crashes on all requests in 1.5-dev12
The haproxy author watch this tag, so will be probably see this. In the meantime, he has instructions for getting a core dump in this mailing list message: permalink.gmane.org/gmane.comp.web.haproxy/7992
Oct
19
comment Will 208v Hurt My Servers?
@sysadmin1138: Happen to have any references for 208 being a little bit more efficient wattage wise than 120?
Oct
12
comment Sender Information Not Cached on the Receiver During an ARP Request
You might also be interested in looking at the tool ettercap. Allows you to play with some ARP attacks (Use it for good of course, and only in lab/home environments).
Sep
25
comment Compare HTTP load balacing using HAProxy & Nginx
serverfault.com/questions/229945/…
Sep
25
comment Should the MAC Tables on a switch Stack be the same between sessions?
@rnxrx: The dynamic count is also different depending on the session it is run from. I also updated my question to show a specific example of a dynamic entry that shows up inside session 3, but not in the "master" session
Sep
20
comment Cluster failover and strange gratuitous arp behavior
Driver is: 9.8.42.0 , teaming is: 17.1.55.0
Sep
17
comment Cluster failover and strange gratuitous arp behavior
Saw the same thing this weekend with Intel ALB, Microsoft failover clustering, and SQL 2012 AlwaysOn on 2008R2. Didn't have a chance to try without teaming. But with Wireshark I observed the backup machine sending out ARPs claiming it still had the the virtual IP Well after the other had machine had taken the VIP over.
Sep
5
comment High Tq values for HAProxy
@Will: We wrote a service that takes syslog data via UDP, parses it with a regex, and inserts it into MS SQL Server. Each day gets its own table.
Sep
5
comment High Tq values for HAProxy
@Will: Okay, will leave this incase someone else has the same question in the future. My guess is a network / resource problem then. Have you checked dmesg and all the vitals? No issues with conntrack table filling etc?
Jul
30
comment HAProxy failing over to a “busy” page
Does this do what you need? serverfault.com/questions/406433/…
Jun
21
comment Tuning Linux + HAProxy
@react: What flags did you make haproxy with?