Tag Info

Hot answers tagged

102

Hitting F1 or h will show you the key. But for reference, the default colors are: CPU: Blue = Low priority threads Green = Normal priority threads Red = Kernel threads Memory: Green = Used memory Blue = Buffers Yellow/Orange = Cache There are a couple of different color-schemes available, you can see them through hitting F2. In your specific case, ...


85

I suggest that you make the adjustments they have requested. Then benchmark the performance to show them that it made no difference. You could even go so far to benchmark it with LESS memory and vCPU to make your point. Also, "We're paying you to support the software with actual solutions, not guesswork."


66

Providing you are confident you are within the given system specs they document. Then any claim they are making in regards to requiring more RAM or CPU they should be able to back up. As the experts in their system I hold people to account on this. Ask them specifics. What information provided on the system indicates more RAM is needed and how did you ...


41

I couldn't find this documented elsewhere. Looking into the code: There are two modes for CPU metrics reporting: the default one, and a "detailed CPU time" which can be enabled from the Setup screen (Display Options / Detailed CPU time). All of them show the percentage of time spent in different processes: Default mode Blue: low priority processes (nice ...


26

Munin data is stored in Round Robin Database files (.rrd). These are stored under /var/lib/munin. Each host and hostgroup will have it's own subdirectory under /var/lib/munin Delete the .rrd files under the hostname in question, and your data will be zeroed out. Graphs will be re-generated after a few minutes.


26

iftop or pktstat -nT (for short term monitoring) is what you need to do this (under *nix). For long-term monitoring, ntop is useful. Finding pktstat is a little tricky for those who aren't running a Debian / Ubuntu box, but this is a decent pktstat source-code archive Use tcpview if you want the same kind of stats under windows


24

Yes, there is an (almost) non-intrusive and easy way: Split each service to run in its own SVCHOST.EXE process and the service consuming the CPU cycles will be easily visible in Process Explorer (the space after "=" is required): SC Config Servicename Type= own Do this in a command line window or put it into a BAT script. Administrative privileges are ...


17

The big thing is to be able to prove that you are using best practices for your system allocation, notably RAM and CPU reservations for your SQL server. All this being said the easiest thing is to make the adjustments requested, at least temporarily. If nothing else it tends to get vendors over feet dragging. I can't count the number of times I've needed ...


16

For this specific situation (where you have VMware and application developers or a third party who does not understand resource allocation), I use a week's worth of metrics obtained from vCenter Operations Manager (vCops - download a demo if needed) to pinpoint the real constraints, bottlenecks and sizing requirements of the application's VM(s). Sometimes, ...


16

For a general sense of the scale of your problem netstat -s will track your total number of retransmissions. # netstat -s | grep retransmited 368644 segments retransmited For a deeper dive, you'll probably want to fire up Wireshark. In Wireshark set your filter to tcp.analysis.retransmission to see retransmissions by flow. That's the best option I ...


14

Check out Apache's LogFormat directive. It allows to log the time taken to serve the request (%D and %T). This can be used for monitoring your server's response time. It will for example tell you if your server responds slower after you have made a change. However, I am not aware of any tool which uses that information to create a report.


14

You want the sysstat utilities: iostat and sar are what you'll find most useful, I think. iostat will give you current stats. sar will grab and store and show you historical stats. Also possibly useful is iotop, which is like iostat but ties things to PIDs. It's newer though, and I don't know as much about it.


13

Per process memory accounting is tricky for a number of reasons I'll get into in a minute. For simple monitoring, gkrellmd, or nagios scripts is probably enough. If you want greater accuracy, you'll need to look harder. smem introduces the concept of Proportional Set Size: Because large portions of physical memory are typically shared among multiple ...


12

Munin is very nice, and easy to install and setup.


12

iostat, sar, top, munin


12

Take a look at: iostat vmstat -d 2 dstat iotop


11

Since you weren't using your memory for anything better, its being put to use as disk cache. If you start services, they will just take memory from the cache. This is a good thing, not a sign that anything is wrong. See linuxatemyram


11

This is a good question because getting a read on memory issues for performance monitoring is difficult. First off, when looking at Page Faults/sec keep in mind that this includes soft faults, hard faults and file cache faults. For the most part, you can ignore soft faults (i.e. paging between memory locations) and cache faults (reading files in to memory) ...


10

Install sysstat package if you don't have it already and then use command sar -d 1. Watch the tps column. It gives you at least the ballpark figure. But remember that actual maximum number of IOPS varies heavily depending on your server workload. A server with lots of sequential access might get more IOPS than a one doing lots of random access.


10

You might also want to have a look at iptraf.


10

You can do it with event handlers. First, add an event handler for your Load average definition: define service{ use generic-service host_name xx service_description Load_Average check_command check_nrpe!check_load event_handler processes_snapshot!xx contact_groups ...


10

Is this a 64bit server - do you have the lock pages in memory local policy enabled? SQL is likely consuming the rest of your memory If you look at the perfmon counters you will see the memory allocation Here is an article that explains it in depth You can also view the counters in SQL SELECT object_name ,Counter_name ,cntr_value ,ROUND(( ...


10

You probably shouldn't consider MyISAM, INNODB will work for you. MyISAM is maybe faster when it comes to SELECT but (for example) it locks your full table on updates. As for INNODB: generally, always consider more RAM before you go into sharding (size of the DB =~ RAM) take a look at the following variables: innodb_buffer_pool_size (we use roughly ...


9

You can't observe what's happening without influencing the object being observed to some degree :-) But the effect is minimal. Otherwise you'd have had a lot of people complaining about how useless performance monitor is when practically used in the field. Edit: The more counters you use, the more overhead it does impose (as from this link). My experiences ...


9

Try the watch command: watch -n 10 ls -l /proc/$$/fd Watch is nice. You could use an old school while loop: while : do ls -l /proc/$$/fd sleep 10 done watch is in the procps package on debian based systems and the procps rpm on RedHat derived systems.


9

I used to work in support - and part of what you're asking sounds highly rational (and probably is): but there are a few questions to ask yourself prior to just doing the "performance enhancement" they're requesting are you running at least at the vendor's stated minimum system requirements already? if you're at least at minimum sysreqs, are you already at ...


8

Either ask to escalate the ticket or ask for a different rep. Depending on which vendor it is escalation may help if you say that you feel that the current level of support doesn't adequately address the issue. If they will not escalate then asking for a different rep may help because that requires much less "justification" since all it needs is to not be ...


8

This is the way I use these words. Others may have additional or different usages. Depending on the job at hand, I will use the terms differently. Development teams and operations teams have different needs an usage. Monitoring is monitoring. Usually it is ongoing, and preferably automated. Open source tools like Munin, Nagios, and MRTG fall into this ...


8

Uhm... iostat on my system shows the IOPS: Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn sda 1.00 64.00 0.00 64 0 Might want to look at upgrading.


8

In general, GigE should be faster than Firewire 800. There are some variables at work that may make FireWire faster, but except for Ethernet congestion from other traffic they're edge cases. If you're looking to connect more than 2 hosts, Ethernet is by far the better choice.



Only top voted, non community-wiki answers of a minimum length are eligible