1,152 reputation
28
bio website deployable.co
location
age
visits member for 2 years, 5 months
seen Oct 30 at 20:06

Jan
13
comment Something eats all memory on my server, but it is NOT fs cache
@Daniil ignore the VIRT column, it doesn't relate to used physical memory.
Jan
13
comment Something eats all memory on my server, but it is NOT fs cache
Can you cat /proc/meminfo and also run ps_mem.py to get a better idea of process memory usage.
Jan
13
comment Apache httpd 2.2.16 creates new sessions every time
Do you have some example request logs from when the issue occurs? Does a request with a JSESSIONID at httpd pass to Tomcat without a JSESSIONID?
Jan
8
answered Remove Corrupt File With Bad File Name Linux
Jan
8
comment Remove Corrupt File With Bad File Name Linux
a double dash ends generally help with file names starting with a - character, so they don't get interpreted as arguments
Jan
7
answered Vagrant single box with multiple addressable domain.locals
Jan
6
comment Use one volume for multiple EC2 servers
@ceejayoz yes, but only for writes of new objects. eveything else is still eventual.
Jan
4
comment hyper-v server time sync source - PDC / external NTP?
plan 3 =)
Jan
4
comment Use one volume for multiple EC2 servers
You can use s3fs via FUSE to make an S3 bucket look like a local filesystem and copy files into it, just remember that even though it looks, walks and quacks like a file system, it's not a real file system and has eventual consistency
Jan
4
answered Use one volume for multiple EC2 servers
Dec
15
answered Why server send RST packet to client after graceful close?
Sep
26
comment How do services with large ingest rates install enough hard disks fast enough?
Not all of them
Sep
26
comment How do services with large ingest rates install enough hard disks fast enough?
Someone still has to build/connect/test the shipping container server at a time and google/facebook tend to do their hardware in house.
Sep
26
comment How do services with large ingest rates install enough hard disks fast enough?
Another facet to think of is drive failure rates with that much storage and how they keep up. Some people have estimated Google might have a new drive failure somewhere every minute.
Sep
26
comment How do services with large ingest rates install enough hard disks fast enough?
The open vault specs are online. With 4TB drives they are up to 120TB in 2U.
Sep
24
answered How do I force apache to drop requests after a predetermined amount of time?
Sep
24
comment How do I force apache to drop requests after a predetermined amount of time?
TCP was certainly made to provide for high latency networks but I'm not sure that precludes it working with low latency requirements. If you control the recv/send ends (and hopefully the network in between) I don't see TCP as the limitation. Most applications and/or stacks can be tuned to prefer low latency/small packets than higher through put that can introduce packet delays. Apache httpd on the other hand...
Sep
20
comment indexing logs for fast repeated grep/search
unix.stackexchange.com/a/66297/22470
Sep
19
comment High load with low CPU usage and low IO usage on Solaris with ZFS and MySQL
Also does the server generally recover on it's own or do you need to restart db's/app servers to clear things out?
Sep
18
comment High load with low CPU usage and low IO usage on Solaris with ZFS and MySQL
can you put up a snapshot of vmstat while the load is increasing?