Skip to main content

Posts

linux - Choosing VPS specs

i'm planning to sign up for a VPS account. On the VPS i'm planning to host ~5 SVN repos, a few php-based lowtraffic websites, a gameserver (MineCraft) and probably some other small apps. The SVN repos will only be updates once or twice a day so they wont take much diskspace or traffic. The websites can be quite rich of images, but has a low ammount of visitors. I don't know wery much about the MineCraft server. But as we will be ~5 players, i don't think it will take much traffic or memory. Now to my question. What specifications do i need? Is 128mb RAM enough? How much traffic/month do i need?

linux - Choosing VPS specs

i'm planning to sign up for a VPS account. On the VPS i'm planning to host ~5 SVN repos, a few php-based lowtraffic websites, a gameserver (MineCraft) and probably some other small apps. The SVN repos will only be updates once or twice a day so they wont take much diskspace or traffic. The websites can be quite rich of images, but has a low ammount of visitors. I don't know wery much about the MineCraft server. But as we will be ~5 players, i don't think it will take much traffic or memory. Now to my question. What specifications do i need? Is 128mb RAM enough? How much traffic/month do i need?

linux - sendmail configuration?

itemprop="text"> I want to send email from a C++ program. I want to do it using sendmail. Well, actually I don't care how, I just found that none of the smtp libaries work... The way to do this, I found via google, is this: echo "This will go into the body of the mail." | mail -s "Subject line" you@youremailid.com The problem is, the mail never arrives at destination. Now what I find peculiar: Shouldn't there be a need for a smtp-server URL? Such as smtp.my-provider.com:25 ? itemprop="text"> class="normal">Answer Sendmail will try to look up the MX record of the domain you're sending to, so you wouldn't have to specify the SMTP server's address or hostname by default. If you do need to force a particular address, that can be han...

linux - sendmail configuration?

I want to send email from a C++ program. I want to do it using sendmail. Well, actually I don't care how, I just found that none of the smtp libaries work... The way to do this, I found via google, is this: echo "This will go into the body of the mail." | mail -s "Subject line" you@youremailid.com The problem is, the mail never arrives at destination. Now what I find peculiar: Shouldn't there be a need for a smtp-server URL? Such as smtp.my-provider.com:25 ? Answer Sendmail will try to look up the MX record of the domain you're sending to, so you wouldn't have to specify the SMTP server's address or hostname by default. If you do need to force a particular address, that can be handled via DNS or an entry in the Sendmail mailertable file. You should first try to determine what happened to your original message. Try sending from your application and running: tail -f /var/log/maillog You can also view the mail queue directly by using t...

hardware - moving raid 10 to another identical server both on Smart Array 6i controllers

itemprop="text"> I have dead HP DL 380G4 with RAID 1+0 with 1 logical volume from 4x72GB drives on built-in Smart Array 6i 128Mb BBWC. It was shut down properly. It seems it was usual death for Proliant with ILO led 2,3,8 lighting. I want to move array to another identical server with same raid firmware level. What is the best strategy?: I have RAID 1+0 on bay 0 bay 1 bay 2 bay 3 As I understand bay0+bay1 are in RAID 1 , bay2+bay3 are in RAID 1, and both RAID 1 pairs are in RAID 0. So should I : Clear RAID config on new server, insert bay 0, bay 2 and power-on or Create RAID 1+0 with 1 logical volume from clear HDDs , and then poweroff ,remove HDDs and insert 2 HDDs(bay 0, bay 2) from old RAID 1+0. then power-on. (each hdd has its raid position info stored but may be could work on same config) Ac...

hardware - moving raid 10 to another identical server both on Smart Array 6i controllers

I have dead HP DL 380G4 with RAID 1+0 with 1 logical volume from 4x72GB drives on built-in Smart Array 6i 128Mb BBWC. It was shut down properly. It seems it was usual death for Proliant with ILO led 2,3,8 lighting. I want to move array to another identical server with same raid firmware level. What is the best strategy?: I have RAID 1+0 on bay 0 bay 1 bay 2 bay 3 As I understand bay0+bay1 are in RAID 1 , bay2+bay3 are in RAID 1, and both RAID 1 pairs are in RAID 0. So should I : Clear RAID config on new server, insert bay 0, bay 2 and power-on or Create RAID 1+0 with 1 logical volume from clear HDDs , and then poweroff ,remove HDDs and insert 2 HDDs(bay 0, bay 2) from old RAID 1+0. then power-on. (each hdd has its raid position info stored but may be could work on same config) According to documentation for Smart Array 6i it could be possible to migrate. however one requirement point is unclear for me Before you move drives, the following conditions must be met: • The array is in its...

virtualization - Best technique for reusing a Windows system image across configurations

itemprop="text"> We are a small company that provides solutions for ventilation systems. Part of the solution is a "controller" which communicates with the ventilation equipment. These controllers are simply Dell computers that come with our Windows 7 system image on them and sometimes some special hardware. We typically do a batch of 10 controllers at a time. We have been using Norton Ghost to apply the system image, but this process breaks because Dell changes the system configuration often, and our Windows image now does not contain the correct drivers. This is especially a problem when they change the RAID controller. To improve this, I see 2 options: use some kind of virtualization and install a hypervisor on each PC. This would solve the driver problem, but probably cause trouble with our special ...

virtualization - Best technique for reusing a Windows system image across configurations

We are a small company that provides solutions for ventilation systems. Part of the solution is a "controller" which communicates with the ventilation equipment. These controllers are simply Dell computers that come with our Windows 7 system image on them and sometimes some special hardware. We typically do a batch of 10 controllers at a time. We have been using Norton Ghost to apply the system image, but this process breaks because Dell changes the system configuration often, and our Windows image now does not contain the correct drivers. This is especially a problem when they change the RAID controller. To improve this, I see 2 options: use some kind of virtualization and install a hypervisor on each PC. This would solve the driver problem, but probably cause trouble with our special hardware. use some method of adding the proper drivers to our Windows image in offline mode. I haven't got much experience in either of these approaches. How would you solve our problem? ...

linux - Why is mongod not using all available RAM?

We have an mongod instance runnning on a VM, and it doesn't seem to be using all available memory. It's page-faulting significantly more than usual, and the system's performance has been significantly degraded lately. More specifically, if I htop mongod , I see: VIRT: 3471G RES: 11.8G The VM has ~60 GB of memory, currently, ~4.6GB is "used", and the remainder is in buffers or cache. My understanding is that mongod mmap s the database files. (This is why VIRT is huge.) However, we're not clear on why the RES number isn't closer to 60 GB: as mongod needs data off disk, this data should be brought into the processes RSS, no? Mongo reports that it is page-faulting, so one would assume that the RSS would grow over time; ours is holding steady. There is nothing else ...

linux - Why is mongod not using all available RAM?

We have an mongod instance runnning on a VM, and it doesn't seem to be using all available memory. It's page-faulting significantly more than usual, and the system's performance has been significantly degraded lately. More specifically, if I htop mongod , I see: VIRT: 3471G RES: 11.8G The VM has ~60 GB of memory, currently, ~4.6GB is "used", and the remainder is in buffers or cache. My understanding is that mongod mmap s the database files. (This is why VIRT is huge.) However, we're not clear on why the RES number isn't closer to 60 GB: as mongod needs data off disk, this data should be brought into the processes RSS, no? Mongo reports that it is page-faulting, so one would assume that the RSS would grow over time; ours is holding steady. There is nothing else significant running on this machine. (It's the database server.) What's consuming the rest of buffers and cache, and specifically, why is the RES size of mongod not expanding to fill ...