SOPHOS CERTIFIED ENGINEER EXAM QUESTIONS WITH VERIFIED
ANSWERS ALREADY GRADED A+
Terms in this set (222)
Required average number of microseconds for a round
300
trip from app server to db (ping) for a 1K block of data
for any supported architecture
9 Milliseconds or less to commit a 1K block of data
17 Milliseconds or less to commit 4K block of data
20 Milliseconds or less to commit a 8K block of data
Recommended. Standard 2 tier webapp arch. Users
Single Facility Architecture
separated from from webapp by firewall. Then they
connect to LB / UI / DB. Task servers also connect to
db.
Recommended. Standard 2 tier web app arch with
Database Architecure w/
SAN supporting the db. Recommended as long as SAN
Storage Area Network (SAN)
is fast eough (under 9 milli to commit 1K data). SAN
need to be in same warehouse as db / app server.
Recommended. There is an active UI / Task / DB and
standby UI / Task DB. Recommended as long as:
Disaster Recovery Support 1. One phase commit (primary can't be blocked until commit
finishes)
2. App servers in secondary not running when primary runs
3. LB only routes to primary
Recommended but not required. In this arch internal
Web Server Layer users go >> firewall >> LB >> web server >> ui >> db.
This configuration is needed if SSO is needed at the
web server level.
Recommended. For this config, internal users go >>
External Access and DMZ firewall >> LB >> ui >> db. External users go >> firewall
Layer
>> LB >> DMZ Reverse Proxy (no ip's exposed to
external users) >> UI >> DB.
Recommended. App database can be in a different data center -
WAN Traversal for Agg w/
use cloud gateway
Cloud Gateway
for max security. Main db still needs to be same data center as
UI / Task servers.
, Not Recommended. 3 tier web architecture. There is a
Dual Firewall firewall between ui / task servers and the database.
This adds to much latency. 1K ping will not return in
under
.3 milli
Not Recommended. Using a shared resource pool
Shared Data Services negatively impacts performance. It also makes it
variable - same ping at different times can take
different amounts of time.
Not Recommended. Ties maintenance windows of IIQ to
Shared App Servers
maintenance windows of other applications.
Competition for resources hurts performance.
Multiple App Servers on Not Recommended. Involves at least one other non
Single (Virtual) Host recommended architecture.
Not Supported. DB server is across WAN, and that
Remote DB
creates unacceptable latency. IIQ needs to be in
same facility as DB.
Remote Shared DB Service Not Supported. WAN creates unacceptable latency and
competition for resources.
Not supported. Primary waits for secondary to finish, so waiting
Two Phase Commit
Replication for WAN hop creates unacceptable latency.
Not supported. Active/Active with real time replication
Multiple Active Sites
relies on WAN hop for all DB writes which results in
unacceptable latency.
Not Supported. 2 data centers - where some app
Remote App Server Nodes
servers are remote. For remote app server WAN hop
for db commit which results in unacceptable latency.
Not supported. All servlet session replication needs to be
Session Replication
turned off ( jboss / tomcat
/ etc)
Less than 5,000 identities. Singled host has app server
and DB. Fine for sandboxes, not for Prod (no
redundancy).
Micro
Specs:
1-2 CPUS
2-4 GB RAM for IIQ Hosts
ANSWERS ALREADY GRADED A+
Terms in this set (222)
Required average number of microseconds for a round
300
trip from app server to db (ping) for a 1K block of data
for any supported architecture
9 Milliseconds or less to commit a 1K block of data
17 Milliseconds or less to commit 4K block of data
20 Milliseconds or less to commit a 8K block of data
Recommended. Standard 2 tier webapp arch. Users
Single Facility Architecture
separated from from webapp by firewall. Then they
connect to LB / UI / DB. Task servers also connect to
db.
Recommended. Standard 2 tier web app arch with
Database Architecure w/
SAN supporting the db. Recommended as long as SAN
Storage Area Network (SAN)
is fast eough (under 9 milli to commit 1K data). SAN
need to be in same warehouse as db / app server.
Recommended. There is an active UI / Task / DB and
standby UI / Task DB. Recommended as long as:
Disaster Recovery Support 1. One phase commit (primary can't be blocked until commit
finishes)
2. App servers in secondary not running when primary runs
3. LB only routes to primary
Recommended but not required. In this arch internal
Web Server Layer users go >> firewall >> LB >> web server >> ui >> db.
This configuration is needed if SSO is needed at the
web server level.
Recommended. For this config, internal users go >>
External Access and DMZ firewall >> LB >> ui >> db. External users go >> firewall
Layer
>> LB >> DMZ Reverse Proxy (no ip's exposed to
external users) >> UI >> DB.
Recommended. App database can be in a different data center -
WAN Traversal for Agg w/
use cloud gateway
Cloud Gateway
for max security. Main db still needs to be same data center as
UI / Task servers.
, Not Recommended. 3 tier web architecture. There is a
Dual Firewall firewall between ui / task servers and the database.
This adds to much latency. 1K ping will not return in
under
.3 milli
Not Recommended. Using a shared resource pool
Shared Data Services negatively impacts performance. It also makes it
variable - same ping at different times can take
different amounts of time.
Not Recommended. Ties maintenance windows of IIQ to
Shared App Servers
maintenance windows of other applications.
Competition for resources hurts performance.
Multiple App Servers on Not Recommended. Involves at least one other non
Single (Virtual) Host recommended architecture.
Not Supported. DB server is across WAN, and that
Remote DB
creates unacceptable latency. IIQ needs to be in
same facility as DB.
Remote Shared DB Service Not Supported. WAN creates unacceptable latency and
competition for resources.
Not supported. Primary waits for secondary to finish, so waiting
Two Phase Commit
Replication for WAN hop creates unacceptable latency.
Not supported. Active/Active with real time replication
Multiple Active Sites
relies on WAN hop for all DB writes which results in
unacceptable latency.
Not Supported. 2 data centers - where some app
Remote App Server Nodes
servers are remote. For remote app server WAN hop
for db commit which results in unacceptable latency.
Not supported. All servlet session replication needs to be
Session Replication
turned off ( jboss / tomcat
/ etc)
Less than 5,000 identities. Singled host has app server
and DB. Fine for sandboxes, not for Prod (no
redundancy).
Micro
Specs:
1-2 CPUS
2-4 GB RAM for IIQ Hosts
