Differences
This shows you the differences between two versions of the page.
Next revision | Previous revision | ||
pandoc:parallel-io:03_storage_infrastructure:storage_infrastructure [2017/12/04 13:11] – Pandoc Auto-commit pandoc | pandoc:parallel-io:03_storage_infrastructure:storage_infrastructure [2020/10/20 09:13] (current) – Pandoc Auto-commit pandoc | ||
---|---|---|---|
Line 16: | Line 16: | ||
* 2x Gigabit Ethernet | * 2x Gigabit Ethernet | ||
* 1x Infiniband QDR | * 1x Infiniband QDR | ||
- | * Login Nodes, Head Nodes, Hypervisors, | + | * Login Nodes, Head Nodes, Hypervisors, |
====== VSC2 - File Systems ====== | ====== VSC2 - File Systems ====== | ||
Line 25: | Line 25: | ||
* Scratch (/ | * Scratch (/ | ||
* TMP (/tmp) | * TMP (/tmp) | ||
- | * Change Filesystem with 'cd' | + | * Change Filesystem with ‘cd’ |
* cd /global # Go into Global | * cd /global # Go into Global | ||
* cd ~ # Go into Home | * cd ~ # Go into Home | ||
* Use your applications settings to set in/out directories | * Use your applications settings to set in/out directories | ||
- | * Pipe the output into your home with './myApp.sh 2>&1 > ~/out.txt' | + | * Pipe the output into your home with ‘./myApp.sh 2>&1 > ~/out.txt’ |
====== VSC - Fileserver ====== | ====== VSC - Fileserver ====== | ||
- | {{pandoc: | + | {{.: |
====== VSC - Why different Filesystems ====== | ====== VSC - Why different Filesystems ====== | ||
Line 45: | Line 45: | ||
* Do some testing | * Do some testing | ||
* Write (small) Log Files | * Write (small) Log Files | ||
- | * etc... | + | * etc… |
====== VSC - Why different Filesystems ====== | ====== VSC - Why different Filesystems ====== | ||
Line 62: | Line 62: | ||
* Small I/Os | * Small I/Os | ||
- | * Random Access | + | * Random Access |
- | * Others | + | * Others |
* Large I/Os | * Large I/Os | ||
- | * Random Access | + | * Random Access |
- | * Sequential Access | + | * Sequential Access |
====== VSC2 - Storage Homes ====== | ====== VSC2 - Storage Homes ====== | ||
Line 112: | Line 112: | ||
* 2x Infiniband QDR (Dual-Rail) | * 2x Infiniband QDR (Dual-Rail) | ||
* 2x Gigabit Ethernet | * 2x Gigabit Ethernet | ||
- | * Login Nodes, Head Nodes, Hypervisors, | + | * Login Nodes, Head Nodes, Hypervisors, |
====== VSC3 - File Systems ====== | ====== VSC3 - File Systems ====== | ||
Line 147: | Line 147: | ||
* 1 Metadata Target per Server | * 1 Metadata Target per Server | ||
* 2 SSDs each (Raid-1 / Mirrored) | * 2 SSDs each (Raid-1 / Mirrored) | ||
- | * Up to 20'000 MB/s throughput | + | * Up to 20’000 MB/s throughput |
* ~ 600 TB capacity | * ~ 600 TB capacity | ||
Line 164: | Line 164: | ||
* Raid (Spectrum Scale Raid) | * Raid (Spectrum Scale Raid) | ||
* Fast rebuild | * Fast rebuild | ||
- | * ... | + | * … |
====== VSC3 - EODC Filesystem ====== | ====== VSC3 - EODC Filesystem ====== | ||
- | * VSC3 <--> EODC | + | * VSC3 <–> EODC |
* 4 IBM ESS Servers (~350 spinning disks each) | * 4 IBM ESS Servers (~350 spinning disks each) | ||
* Running GPFS 4.2.3 | * Running GPFS 4.2.3 | ||
Line 184: | Line 184: | ||
====== VSC3 - EODC Filesystem ====== | ====== VSC3 - EODC Filesystem ====== | ||
- | * VSC3 is a "remote cluster" | + | * VSC3 is a “remote cluster” for the EODC GPFS filesystem |
* 2 Management servers | * 2 Management servers | ||
* Tie-Breaker disk for quorum | * Tie-Breaker disk for quorum | ||
Line 192: | Line 192: | ||
====== VSC3 - Bioinformatics ====== | ====== VSC3 - Bioinformatics ====== | ||
- | * VSC3 got a "bioinformatics" | + | * VSC3 got a “bioinformatics” upgrade in late 2016 |
* 17 Nodes | * 17 Nodes | ||
* 2x Intel Xeon E5-2690 v4 @ 2.60 Ghz (14 cores each / 28 with hyperthreading) | * 2x Intel Xeon E5-2690 v4 @ 2.60 Ghz (14 cores each / 28 with hyperthreading) | ||
Line 204: | Line 204: | ||
====== VSC3 - Bioinformatics ====== | ====== VSC3 - Bioinformatics ====== | ||
- | {{pandoc: | + | {{.:binf.jpg}} |
====== VSC3 - BINFL Filesystem ====== | ====== VSC3 - BINFL Filesystem ====== | ||
Line 249: | Line 249: | ||
====== Storage Performance ====== | ====== Storage Performance ====== | ||
- | {{pandoc: | + | {{.: |
====== Temporary Filesystems ====== | ====== Temporary Filesystems ====== | ||
Line 273: | Line 273: | ||
* via self written, parallellized rsync script | * via self written, parallellized rsync script | ||
* Global | * Global | ||
- | * Metadata is backupped. But if a total disaster happens this won't help. | + | * Metadata is backupped. But if a total disaster happens this won’t help. |
* Altough nothing happenend for some time, this is high performance computing and redundancy is minimal | * Altough nothing happenend for some time, this is high performance computing and redundancy is minimal | ||
- | * Keep an eye on your data. If it's important you should backup it yourself. | + | * Keep an eye on your data. If it’s important you should backup it yourself. |
* Use rsync | * Use rsync | ||
====== Addendum: Big Files ====== | ====== Addendum: Big Files ====== | ||
- | * What does an administrator mean with 'don't use small files'? | + | * What does an administrator mean with ‘don’t use small files’? |
* It depends | * It depends | ||
- | * On /global fopen --> fclose takes ~100 microseconds | + | * On /global fopen –> fclose takes ~100 microseconds |
* On VSC2 Home Filesystems > 100 Million Files are stored. | * On VSC2 Home Filesystems > 100 Million Files are stored. | ||
* A check which files have changed takes more than 12 hours. Without even reading file contents or copying. | * A check which files have changed takes more than 12 hours. Without even reading file contents or copying. | ||
Line 292: | Line 292: | ||
* Storage works well when the blocksize is reasonable for the storage system (on VSC3 a few Megabytes are enough) | * Storage works well when the blocksize is reasonable for the storage system (on VSC3 a few Megabytes are enough) | ||
* Do not create millions of files | * Do not create millions of files | ||
- | * If every user had millions of files we'd run into some problems | + | * If every user had millions of files we’d run into some problems |
- | * Use 'tar' | + | * Use ‘tar’ to archive unneeded files |
* tar -cvpf myArchive.tar myFolderWithFiles/ | * tar -cvpf myArchive.tar myFolderWithFiles/ | ||
* tar -cvjpf myArchive.tar myFolderWithFiles/ | * tar -cvjpf myArchive.tar myFolderWithFiles/ |