twblamer's questions

For reasons above my paygrade, I need to run multiple instances of a DB server on a physical host with a bare metal OS. I also need to limit the physical memory use of each instance.

I am using a template unit file to start up a few dozen instances. I am using EnvironmentFile to specify different environment variables and command line arguments for each instance.

However, I also want to be able to specify a different memory limit for each instance. AFAICT, environment variable substitution does not work for resource limit options like MemoryHigh.

I thought drop-in files would be the answer, but I couldn't a single example of using drop-in files with an instantiated service. Is there some way I can create one drop-in file for myservice@1, a different drop-in file for myservice@2, etc?

$ cat /etc/systemd/system/[email protected]

[Unit]
Description=Something
After=syslog.target network.target

[Service]
Type=forking    
EnvironmentFile=/etc/myservice/myservice.%i.env
ExecStart=/usr/bin/myservice $OPTIONS_FROM_ENV
MemoryHigh=4G

[Install]
WantedBy=multi-user.target

For example, is there a tool or method where I can say (for example) "generate up to 50 MB/s of read activity to this file, but no greater"? Or configure a tiny delay between read operations, which could have basically the same effect? I have a specific need to create an I/O workload that's more deterministic than a normal dd or disk stress test. I'm looking for something that works in Linux...

Let's say I know the drive I want is on Adapter 8, Virtual Drive 0, is there a way to get the Linux device name (e.g. /dev/sdt)? I am on RHEL 6.2.

-pdlistdisplays a WWN value for each physical drive, but that doesn't help much. -ldinfo doesn't print a WWN or any other identifying value.

# ./MegaCli64 -pdlist -a8 | grep WWN
WWN: 500051610003776C
WWN: 5000516100037BFC
WWN: 5000516100038090
WWN: 50005161000284F4
WWN: 5000516100037C0C
WWN: 5000516100037C5C
WWN: 50005161000310C8
WWN: 5000516100037F4C

# ls -1dt /dev/disk/by-id/wwn*
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd142c12c84d
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd142c0973e6
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd132c002573
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd122bf6e2f6
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd122beda52a
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd112be46f10
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd112bdb3ffb
/dev/disk/by-id/wwn-0x600605b0043e52d01721fd102bd20f14

# ./MegaCli64 -v                       
 MegaCLI SAS RAID Management Tool  Ver 8.03.08 Mar 15, 2012

Title pretty much explains it. Let's say there's 100 servers, various versions of Windows and Linux, and one Windows server is the "master clock." I did look at this question:

How do I synchronize clocks between Linux and Windows?

This hints that ntp can do what I want if I run "ntpd -q" on a client (?). If I install ntp I also need to guarantee that it will only sync the times when I force it to. Even better if I have a log that tells me every time a sync was performed.

I'm doing benchmark runs and I need to be able to say something like this: "Clocks were synced on all the benchmark systems at 09:42:01am on the master. A benchmark run was then initiated and allowed to run for six hours. None of the system clocks were altered during this time interval."

I understand there is subsequent clock drift, but for now that's the way we're doing things and I'm doing it with a manual process. I'd rather at least automate the one-time sync.